Task description

The goal of acoustic scene classification is to classify a test recording into one of the predefined ten acoustic scene classes. This targets acoustic scene classification with devices with low computational and memory allowance, which impose certain limits on the model complexity, such as the model’s number of parameters and the multiply-accumulate operations count. In addition to low-complexity, the aim is generalization across a number of different devices. For this purpose, the task will use audio data recorded and simulated with a variety of devices.

The development dataset consists of recordings from 10 European cities using 9 different devices: 3 real devices (A, B, C) and 6 simulated devices (S1-S6). Data from devices B, C, and S1-S6 consists of randomly selected segments from the simultaneous recordings, therefore all overlap with the data from device A, but not necessarily with each other. The total amount of audio in the development set is 64 hours.

The evaluation dataset contains data from 12 cities, 10 acoustic scenes, 11 devices. There are five new devices (not available in the development set): real device D and simulated devices S7-S11. Evaluation data contains 22 hours of audio.

The device A consists in a Soundman OKM II Klassik/studio A3, electret binaural microphone and a Zoom F8 audio recorder using 48kHz sampling rate and 24-bit resolution. The other devices are commonly available customer devices: device B is a Samsung Galaxy S7, device C is iPhone SE, and device D is a GoPro Hero5 Session.

More detailed task description can be found in the task description page

Systems ranking

	Submission information								Evaluation dataset		Development dataset
Rank	Submission label	Name	Technical Report	Official system rank	Rank value	Performance rank	Memory rank	MACs rank	Accuracy with 95% confidence interval (Evaluation dataset)	Logloss with 95% confidence interval (Evaluation dataset)	Accuracy (Development dataset)	Logloss (Development dataset)
	AI4EDGE_IPL_task1_1	AI4EDGE_1	Almeida2023	28	30.25	34	14	39	51.9 (51.7 - 52.2)	1.920 (1.786 - 2.054)	67.5	0.932
	AI4EDGE_IPL_task1_2	AI4EDGE_2	Almeida2023	47	43.75	51	25	48	48.8 (48.5 - 49.1)	1.996 (1.859 - 2.133)	69.9	0.872
	AI4EDGE_IPL_task1_3	AI4EDGE_3	Almeida2023	38	37.25	43	23	40	50.8 (50.5 - 51.1)	1.364 (1.308 - 1.420)	66.0	0.969
	Bai_JLESS_task1_1	JLESS	Du2023	54	50.25	56	48	41	47.9 (47.6 - 48.2)	1.825 (1.727 - 1.923)	50.8	1.436
	Bai_JLESS_task1_2	JLESS	Du2023	47	43.75	52	47	24	48.8 (48.5 - 49.1)	1.791 (1.691 - 1.891)	50.5	1.467
	Cai_TENCENT_task1_1	Cai_1	Cai2023	17	25.75	7	45	44	56.6 (56.3 - 56.9)	1.174 (1.122 - 1.225)	57.5	1.147
	Cai_TENCENT_task1_2	Cai_2	Cai2023	11	21.25	10	31	34	56.2 (55.9 - 56.5)	1.246 (1.196 - 1.295)	58.1	1.178
	Cai_TENCENT_task1_3	Cai_3	Cai2023	14	22.25	12	31	34	55.8 (55.5 - 56.1)	1.241 (1.194 - 1.288)	57.4	1.190
	Cai_TENCENT_task1_4	Cai_4	Cai2023	11	21.25	17	21	30	55.4 (55.1 - 55.7)	1.252 (1.207 - 1.297)	57.0	1.198
	Cai_XJTLU_task1_1	TFSepNet1	Cai2023a	9	19.75	36	2	5	51.9 (51.6 - 52.2)	1.307 (1.261 - 1.352)	53.9	1.257
	Cai_XJTLU_task1_2	TFSepNet2	Cai2023a	8	18.25	33	2	5	52.5 (52.3 - 52.8)	1.292 (1.246 - 1.337)	51.9	1.307
	Cai_XJTLU_task1_3	TFSepNet3	Cai2023a	6	14.00	22	5	7	55.1 (54.8 - 55.4)	1.223 (1.173 - 1.273)	57.5	1.160
	Cai_XJTLU_task1_4	TFSepNet4	Cai2023a	3	11.50	5	18	18	57.0 (56.7 - 57.3)	1.241 (1.184 - 1.299)	64.3	0.989
	Fei_vv_task1_1	vv_1	Fei2023	15	24.75	19	38	23	55.2 (55.0 - 55.5)	1.282 (1.223 - 1.341)	59.3	1.145
	Fei_vv_task1_2	vv_2	Fei2023	13	21.75	24	26	13	54.5 (54.2 - 54.8)	1.290 (1.232 - 1.348)	56.7	1.204
	Fei_vv_task1_3	vv_3	Fei2023	26	29.25	28	38	23	53.2 (53.0 - 53.5)	1.349 (1.286 - 1.412)	58.0	1.231
	Fei_vv_task1_4	vv_4	Fei2023	23	28.25	37	26	13	51.8 (51.5 - 52.0)	1.370 (1.306 - 1.434)	55.4	1.284
	Han_SZU_task1_1	Han_SZU_1	Han2023	39	37.75	44	13	50	50.5 (50.3 - 50.8)	2.011 (1.886 - 2.137)	51.4	1.378
	LAM_AEV_task1_1	AEV_sys_1	Pham2023	25	29.00	18	33	47	55.3 (55.1 - 55.6)	1.847 (1.726 - 1.968)	56.8	1.349
	LAM_AEV_task1_2	AEV_sys_2	Pham2023	31	31.50	23	33	47	55.0 (54.7 - 55.3)	2.083 (1.941 - 2.224)	57.4	1.333
	LAM_AEV_task1_3	AEV_sys_3	Pham2023	20	27.00	14	33	47	55.6 (55.3 - 55.9)	1.933 (1.810 - 2.055)	57.4	1.333
	Liang_NTES_task1_1	NTES_1	Liang2023	40	38.50	32	41	49	52.6 (52.3 - 52.8)	1.402 (1.339 - 1.465)	54.9	1.293
	MALACH23_JKU_task1_1	RFR-CNN-1	Pichler2023	8	18.25	6	36	25	57.0 (56.7 - 57.3)	1.230 (1.180 - 1.279)	55.2	1.280
	MALACH23_JKU_task1_2	RFR-CNN-2	Pichler2023	7	16.75	8	32	19	56.6 (56.3 - 56.8)	1.242 (1.196 - 1.288)	53.5	1.323
	MALACH23_JKU_task1_3	S4-1	Pichler2023	42	39.50	67	22	2	9.9 (9.7 - 10.1)	4.354 (4.289 - 4.418)	46.7	1.496
	MALACH23_JKU_task1_4	S4-2	Pichler2023	43	39.75	68	22	1	9.8 (9.7 - 10.0)	3.224 (3.184 - 3.265)	45.1	1.509
	DCASE2023 baseline	Baseline		52	46.75	64	13	46	44.8 (44.5 - 45.1)	1.523 (1.478 - 1.568)	42.9	1.575
	Park_KT_task1_1	KT_1	Kim2023	10	20.75	9	34	31	56.3 (56.0 - 56.6)	1.495 (1.410 - 1.580)	72.5	0.824
	Park_KT_task1_2	KT_2	Kim2023	29	30.75	29	34	31	53.0 (52.7 - 53.3)	1.660 (1.569 - 1.751)	56.1	1.446
	Park_KT_task1_3	KT_3	Kim2023	26	29.25	26	34	31	54.2 (54.0 - 54.5)	2.230 (2.107 - 2.353)	70.5	1.167
	Park_KT_task1_4	KT_MF	Kim2023	20	27.00	49	7	3	49.2 (48.9 - 49.5)	1.510 (1.469 - 1.550)	54.2	1.427
	Schmid_CPJKU_task1_1	CPM bc=8	Schmid2023	5	13.75	25	1	4	54.4 (54.1 - 54.6)	1.313 (1.245 - 1.380)	52.6	1.370
	Schmid_CPJKU_task1_2	CPM bc=16	Schmid2023	1	5.25	4	4	9	58.7 (58.4 - 59.0)	1.256 (1.181 - 1.332)	58.4	1.200
	Schmid_CPJKU_task1_3	CPM bc=24	Schmid2023	2	7.00	2	8	16	61.4 (61.2 - 61.7)	1.153 (1.085 - 1.221)	61.8	1.090
	Schmid_CPJKU_task1_4	CPM bc=32	Schmid2023	3	11.50	1	16	28	62.7 (62.4 - 63.0)	1.117 (1.047 - 1.187)	64.1	1.007
	Schmidt_FAU_task1_1	30mmacs	Schmidt2023	26	29.25	13	46	45	55.7 (55.4 - 56.0)	1.322 (1.271 - 1.373)	57.5	1.217
	Schmidt_FAU_task1_2	20mmacs	Schmidt2023	12	21.50	15	24	32	55.6 (55.3 - 55.9)	1.337 (1.281 - 1.392)	57.2	1.211
	Schmidt_FAU_task1_3	10mmacs	Schmidt2023	19	26.75	31	28	17	52.7 (52.4 - 53.0)	1.398 (1.346 - 1.450)	54.5	1.298
	Schmidt_FAU_task1_4	5mmacs	Schmidt2023	24	28.75	48	9	10	49.7 (49.4 - 50.0)	1.482 (1.427 - 1.536)	50.6	1.417
	Tan_NTU_task1_1	TYPG_T1_1	Tan2023	33	33.50	59	10	6	47.1 (46.8 - 47.4)	1.508 (1.461 - 1.554)	50.3	1.397
	Tan_NTU_task1_2	TYPG_T1_2	Tan2023	35	34.00	54	17	11	48.5 (48.2 - 48.8)	1.461 (1.417 - 1.505)	52.1	1.372
	Tan_NTU_task1_3	TYPG_T1_3	Tan2023	37	37.00	60	17	11	46.3 (46.1 - 46.6)	1.492 (1.449 - 1.534)	50.0	1.381
	Tan_SCUT_task1_1	BSConv1_1	Tan2023a	4	13.50	3	27	21	60.8 (60.6 - 61.1)	1.192 (1.119 - 1.265)	55.7	1.318
	Tan_SCUT_task1_2	BSConv1_2	Tan2023a	30	31.00	38	27	21	51.7 (51.4 - 52.0)	1.444 (1.378 - 1.509)	54.3	1.243
	Tan_SCUT_task1_3	BSConv1_3	Tan2023a	16	25.50	27	27	21	53.5 (53.2 - 53.8)	1.441 (1.370 - 1.513)	55.5	1.320
	Tan_SCUT_task1_4	BSConv1_4	Tan2023a	32	33.00	42	27	21	50.9 (50.6 - 51.2)	1.525 (1.455 - 1.595)	54.5	1.305
	Vo_DU_task1_1	HKD-MLA-1	Vo2023	52	46.75	63	35	26	45.0 (44.7 - 45.2)	2.157 (2.035 - 2.279)	46.0	1.591
	Vo_DU_task1_2	HKD-MLA-2	Vo2023	53	47.75	65	35	26	44.8 (44.5 - 45.1)	2.116 (2.003 - 2.229)	45.4	1.640
	Vo_DU_task1_3	HKD-MLA-3	Vo2023	50	46.25	62	35	26	45.2 (44.9 - 45.5)	2.092 (1.973 - 2.211)	45.8	1.624
	Vo_DU_task1_4	HKD-MLA-4	Vo2023	49	45.75	61	35	26	45.5 (45.2 - 45.8)	1.793 (1.717 - 1.869)	46.7	1.617
	Wang_SCUT_task1_1	DSSDM1	Wang2023	31	31.50	50	11	15	49.1 (48.9 - 49.4)	1.493 (1.434 - 1.553)	53.3	1.287
	Wang_SCUT_task1_2	DSSDM2	Wang2023	18	26.50	30	19	27	52.9 (52.6 - 53.2)	1.348 (1.300 - 1.397)	56.4	1.191
	Wang_SCUT_task1_3	DSSDM3	Wang2023	46	43.50	58	20	38	47.1 (46.8 - 47.4)	1.702 (1.621 - 1.782)	50.8	1.477
	Wang_SCUT_task1_4	DSSDM4	Wang2023	48	45.00	55	37	33	48.5 (48.2 - 48.8)	1.472 (1.416 - 1.529)	52.4	1.368
	XuQianHu_BIT&NUDT_task1_1	DYXS_t1_1	Yu2023	45	40.50	41	44	36	51.0 (50.7 - 51.3)	1.364 (1.319 - 1.409)	59.0	1.164
	XuQianHu_BIT&NUDT_task1_2	DYXS_t1_2	Yu2023	44	40.00	39	40	42	51.6 (51.3 - 51.9)	1.355 (1.308 - 1.401)	60.6	1.141
	XuQianHu_BIT&NUDT_task1_3	DYXS_t1_3	Yu2023	41	39.25	46	43	22	50.0 (49.8 - 50.3)	1.395 (1.346 - 1.445)	59.6	1.168
	XuQianHu_BIT&NUDT_task1_4	DYXS_t1_4	Yu2023	36	35.50	40	42	20	51.1 (50.9 - 51.4)	1.367 (1.324 - 1.411)	61.3	1.139
	Yang_GZHU_task1_1	dml_kd	Weng2023	15	24.75	16	30	37	55.5 (55.3 - 55.8)	1.280 (1.220 - 1.339)	59.7	1.151
	Yang_GZHU_task1_2	dml_kd_tta	Weng2023	14	22.25	11	30	37	55.9 (55.6 - 56.2)	1.241 (1.184 - 1.298)	59.9	1.115
	Yang_GZHU_task1_3	dml	Weng2023	21	27.25	21	30	37	55.1 (54.8 - 55.4)	1.279 (1.217 - 1.340)	57.9	1.170
	Yang_GZHU_task1_4	dml_tta	Weng2023	19	26.75	20	30	37	55.2 (54.9 - 55.5)	1.259 (1.201 - 1.318)	58.0	1.163
	Zhang_NCUT_task1_1	Zhang1_NCUT	Zhang2023	49	45.75	66	39	12	43.3 (43.0 - 43.6)	1.757 (1.692 - 1.821)	47.0	1.671
	Zhang_NCUT_task1_2	Zhang2_NCUT	Zhang2023	51	46.50	57	29	43	47.9 (47.6 - 48.2)	1.533 (1.476 - 1.590)	52.8	1.347
	Zhang_SATLab_task1_1	SATLab_1	Bing2023	26	29.25	53	3	8	48.8 (48.5 - 49.1)	3.248 (3.031 - 3.465)	65.4	0.941
	Zhang_SATLab_task1_2	SATLab_2	Bing2023	34	33.75	47	12	29	50.0 (49.7 - 50.2)	4.213 (3.928 - 4.497)	54.5	1.437
	Zhang_SATLab_task1_3	SATLab_3	Bing2023	27	30.00	35	15	35	51.9 (51.6 - 52.2)	1.704 (1.620 - 1.789)	57.0	1.342
	Zhang_SATLab_task1_4	SATLab_4	Bing2023	22	27.50	45	6	14	50.3 (50.0 - 50.6)	1.542 (1.483 - 1.601)	53.2	1.439

Teams ranking

	Submission information								Evaluation dataset		Development dataset
Rank	Submission label	Name	Technical Report	Official team rank	Rank value	Performance rank	Memory rank	MACs rank	Accuracy with 95% confidence interval (Evaluation dataset)	Logloss with 95% confidence interval (Evaluation dataset)	Accuracy (Development dataset)	Logloss (Development dataset)
	AI4EDGE_IPL_task1_1	AI4EDGE_1	Almeida2023	13	30.25	34	14	39	51.9 (51.7 - 52.2)	1.920 (1.786 - 2.054)	67.5	0.932
	Bai_JLESS_task1_2	JLESS	Du2023	18	43.75	52	47	24	48.8 (48.5 - 49.1)	1.791 (1.691 - 1.891)	50.5	1.467
	Cai_TENCENT_task1_2	Cai_2	Cai2023	6	21.25	10	31	34	56.2 (55.9 - 56.5)	1.246 (1.196 - 1.295)	58.1	1.178
	Cai_XJTLU_task1_4	TFSepNet4	Cai2023a	2	11.50	5	18	18	57.0 (56.7 - 57.3)	1.241 (1.184 - 1.299)	64.3	0.989
	Fei_vv_task1_2	vv_2	Fei2023	8	21.75	24	26	13	54.5 (54.2 - 54.8)	1.290 (1.232 - 1.348)	56.7	1.204
	Han_SZU_task1_1	Han_SZU_1	Han2023	16	37.75	44	13	50	50.5 (50.3 - 50.8)	2.011 (1.886 - 2.137)	51.4	1.378
	LAM_AEV_task1_3	AEV_sys_3	Pham2023	11	27.00	14	33	47	55.6 (55.3 - 55.9)	1.933 (1.810 - 2.055)	57.4	1.333
	Liang_NTES_task1_1	NTES_1	Liang2023	17	38.50	32	41	49	52.6 (52.3 - 52.8)	1.402 (1.339 - 1.465)	54.9	1.293
	MALACH23_JKU_task1_2	RFR-CNN-2	Pichler2023	4	16.75	8	32	19	56.6 (56.3 - 56.8)	1.242 (1.196 - 1.288)	53.5	1.323
	DCASE2023 baseline	Baseline		21	46.75	64	13	46	44.8 (44.5 - 45.1)	1.523 (1.478 - 1.568)	42.9	1.575
	Park_KT_task1_1	KT_1	Kim2023	5	20.75	9	34	31	56.3 (56.0 - 56.6)	1.495 (1.410 - 1.580)	72.5	0.824
	Schmid_CPJKU_task1_2	CPM bc=16	Schmid2023	1	5.25	4	4	9	58.7 (58.4 - 59.0)	1.256 (1.181 - 1.332)	58.4	1.200
	Schmidt_FAU_task1_2	20mmacs	Schmidt2023	7	21.50	15	24	32	55.6 (55.3 - 55.9)	1.337 (1.281 - 1.392)	57.2	1.211
	Tan_NTU_task1_1	TYPG_T1_1	Tan2023	14	33.50	59	10	6	47.1 (46.8 - 47.4)	1.508 (1.461 - 1.554)	50.3	1.397
	Tan_SCUT_task1_1	BSConv1_1	Tan2023a	3	13.50	3	27	21	60.8 (60.6 - 61.1)	1.192 (1.119 - 1.265)	55.7	1.318
	Vo_DU_task1_4	HKD-MLA-4	Vo2023	19	45.75	61	35	26	45.5 (45.2 - 45.8)	1.793 (1.717 - 1.869)	46.7	1.617
	Wang_SCUT_task1_2	DSSDM2	Wang2023	10	26.50	30	19	27	52.9 (52.6 - 53.2)	1.348 (1.300 - 1.397)	56.4	1.191
	XuQianHu_BIT&NUDT_task1_4	DYXS_t1_4	Yu2023	15	35.50	40	42	20	51.1 (50.9 - 51.4)	1.367 (1.324 - 1.411)	61.3	1.139
	Yang_GZHU_task1_2	dml_kd_tta	Weng2023	9	22.25	11	30	37	55.9 (55.6 - 56.2)	1.241 (1.184 - 1.298)	59.9	1.115
	Zhang_NCUT_task1_1	Zhang1_NCUT	Zhang2023	20	45.75	66	39	12	43.3 (43.0 - 43.6)	1.757 (1.692 - 1.821)	47.0	1.671
	Zhang_SATLab_task1_4	SATLab_4	Bing2023	12	27.50	45	6	14	50.3 (50.0 - 50.6)	1.542 (1.483 - 1.601)	53.2	1.439

System complexity

	Submission information		Evaluation dataset				Acoustic model					System
Rank	Submission label	Technical Report	Official system rank	Rank value	Accuracy (Eval)	Logloss (Eval)	MACS	Memory use	Parameters	Non-zero parameters	Sparsity	Complexity management
	AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	1.920	25475456	62720	52852	51986	0.016385378036781972	weight quantization
	AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	1.996	29304736	53760	68996	67826	0.016957504782885935	weight quantization
	AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	1.364	26711936	49280	65192	64034	0.017762915695177295	weight quantization, pruning
	Bai_JLESS_task1_1	Du2023	54	50.25	47.9	1.825	27931612	78252	78252	78252	0.0	model compression
	Bai_JLESS_task1_2	Du2023	47	43.75	48.8	1.791	14130372	60458	60458	60458	0.0	model compression
	Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	1.174	28840396	127684	127684	127684	0.0	weight quantization
	Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	1.246	21990724	79942	79942	79942	0.0	weight quantization
	Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	1.241	21990724	79942	79942	79942	0.0	weight quantization
	Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	1.252	19533124	63558	63558	63558	0.0	weight quantization
	Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	1.307	1649349	6828	6828	6828	0.0	knowledge distillation, weight quantization
	Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	1.292	1649349	6828	6828	6828	0.0	weight quantization
	Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	1.223	3424245	15890	15890	15890	0.0	weight quantization
	Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	1.241	10219540	54260	54260	54260	0.0	weight quantization
	Fei_vv_task1_1	Fei2023	15	24.75	55.2	1.282	13402932	123636	123636	123636	0.0	weight quantization
	Fei_vv_task1_2	Fei2023	13	21.75	54.5	1.290	7802348	70588	70588	70588	0.0	weight quantization
	Fei_vv_task1_3	Fei2023	26	29.25	53.2	1.349	13402932	123636	123636	123636	0.0	weight quantization
	Fei_vv_task1_4	Fei2023	23	28.25	51.8	1.370	7802348	70588	70588	70588	0.0	weight quantization
	Han_SZU_task1_1	Han2023	39	37.75	50.5	2.012	29.349M	80845	80845	80845	0.0	knowledge distillation
	LAM_AEV_task1_1	Pham2023	25	29.00	55.3	1.847	29267550	88704	22962	22176	0.034230467729291836
	LAM_AEV_task1_2	Pham2023	31	31.50	55.0	2.083	29267550	88704	22962	22176	0.034230467729291836
	LAM_AEV_task1_3	Pham2023	20	27.00	55.6	1.933	29267550	88704	22962	22176	0.034230467729291836
	Liang_NTES_task1_1	Liang2023	40	38.50	52.6	1.402	29591778	143345	31260	31260	0.0
	MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	1.230	14686940	119608	59804	59804	0.0
	MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	1.242	10819292	87160	43580	43580	0.0	efficient models
	MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	4.354	572340	116008	116648	29162	0.75	efficient models
	MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	3.224	214420	63592	15994	15994	0.0	efficient models
	DCASE2023 baseline		52	46.75	44.8	1.523	29234920	65280	46512	46512	0.0	weight quantization
	Park_KT_task1_1	Kim2023	10	20.75	56.3	1.495	19556096	250000	92070	92070	0.0	weight quantization
	Park_KT_task1_2	Kim2023	29	30.75	53.0	1.660	19556096	250000	92070	92070	0.0	weight quantization
	Park_KT_task1_3	Kim2023	26	29.25	54.2	2.230	19556096	250000	92070	92070	0.0	weight quantization
	Park_KT_task1_4	Kim2023	20	27.00	49.2	1.510	617000	30000	20516	20516	0.0	weight quantization
	Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	1.313	1582336	5722	5722	5722	0.0	knowledge distillation, weight quantization
	Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	1.256	4354304	12310	12310	12310	0.0	knowledge distillation, weight quantization
	Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	1.153	9638144	30106	30106	30106	0.0	knowledge distillation, weight quantization
	Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	1.117	16803072	54182	54182	54182	0.0	knowledge distillation, weight quantization, structured pruning
	Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	1.322	28931380	127988	127988	127988	0.0	weight quantization, pruning
	Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	1.337	19910080	68456	68456	68456	0.0	weight quantization, pruning
	Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	1.398	9996775	74700	74700	74700	0.0	weight quantization, pruning
	Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	1.482	4938255	34616	34616	34616	0.0	weight quantization, pruning
	Tan_NTU_task1_1	Tan2023	33	33.50	47.1	1.508	2960384	40960	37434	37306	0.0034193513917828433	weight quantization
	Tan_NTU_task1_2	Tan2023	35	34.00	48.5	1.461	6462656	54304	54242	54098	0.002654769366911225	weight quantization
	Tan_NTU_task1_3	Tan2023	37	37.00	46.3	1.492	6462656	54304	54242	54098	0.002654769366911225	weight quantization
	Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	1.192	13180000	73386	73386	73386	0.0	weight quantization
	Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	1.444	13180000	73386	73386	73386	0.0	weight quantization
	Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	1.441	13180000	73386	73386	73386	0.0	weight quantization
	Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	1.525	13180000	73386	73386	73386	0.0	weight quantization
	Vo_DU_task1_1	Vo2023	52	46.75	45.0	2.157	15600000	503316	119526	119526	0.0	weight quantization
	Vo_DU_task1_2	Vo2023	53	47.75	44.8	2.116	15600000	503316	119526	119526	0.0	weight quantization
	Vo_DU_task1_3	Vo2023	50	46.25	45.2	2.092	15600000	503316	119526	119526	0.0	weight quantization
	Vo_DU_task1_4	Vo2023	49	45.75	45.5	1.793	15600000	503316	119526	119526	0.0	weight quantization
	Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	1.493	8646000	62080	45164	45164	0.0	weight quantization
	Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	1.348	16746000	81150	56172	56172	0.0	weight quantization
	Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	1.702	25442000	82280	56556	56556	0.0	weight quantization
	Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	1.472	20902000	148618	121812	121812	0.0	weight quantization
	XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	1.364	23803968	125885	52288	52288	0.0	weight quantization
	XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	1.355	28400320	123654	51648	51648	0.0	weight quantization
	XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	1.395	13402688	125650	57392	57392	0.0	weight quantization
	XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	1.367	11878580	125057	66114	66114	0.0	weight quantization
	Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	1.280	23970000	76906	76906	76906	0.0	weight quantization
	Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	1.241	23970000	76906	76906	76906	0.0	weight quantization
	Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	1.279	23970000	76906	76906	76906	0.0	weight quantization
	Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	1.259	23970000	76906	76906	76906	0.0	weight quantization
	Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	1.757	7375000	574464	123648	123648	0.0	weight quantization
	Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	1.533	28461000	1622016	76224	76224	0.0	weight quantization
	Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	3.248	3972096	7946	7946	7434	0.06443493581676318	weight quantization
	Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	4.213	19466240	46232	46232	45996	0.005104689392628536	pruning, weight quantization
	Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	1.704	23438336	54178	54178	53430	0.013806342057661736	pruning, weight quantization
	Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	1.542	7944192	15892	15892	14868	0.06443493581676318	weight quantization

Energy Consumption

	Submission information		Evaluation dataset			Acoustic model		Normalized energy consumption		Energy consumption (kWh)
Rank	Submission label	Technical Report	Official system rank	Rank value	Accuracy (Eval)	MACS	Memory use	Training	Inference	Training	Inference
	AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	25475456	62720	0.7283	0.7906	0.2540	0.2340
	AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	29304736	53760	0.6146	0.6401	0.3010	0.2890
	AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	26711936	49280	0.6336	0.6655	0.2920	0.2780
	Bai_JLESS_task1_1	Du2023	54	50.25	47.9	27931612	78252	2.5391	2.3548	0.1150	0.1240
	Bai_JLESS_task1_2	Du2023	47	43.75	48.8	14130372	60458	2.5614	2.4132	0.1140	0.1210
	Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	28840396	127684	0.0962	0.2854	0.8510	0.2870
	Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	21990724	79942	0.1091	0.2925	0.7510	0.2800
	Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	21990724	79942	0.1091	0.2925	0.7510	0.2800
	Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	19533124	63558	0.1185	0.3302	0.6910	0.2480
	Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	1649349	6828	0.0010	0.2500	2.0300	0.0080
	Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	1649349	6828	0.0013	0.2500	1.5910	0.0080
	Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	3424245	15890	0.0012	0.1250	1.7280	0.0160
	Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	10219540	54260	0.0011	0.0952	1.7550	0.0210
	Fei_vv_task1_1	Fei2023	15	24.75	55.2	13402932	123636	0.6742	8.0000	1.7800	0.1500
	Fei_vv_task1_2	Fei2023	13	21.75	54.5	7802348	70588	0.7595	8.0000	1.5800	0.1500
	Fei_vv_task1_3	Fei2023	26	29.25	53.2	13402932	123636	0.6742	8.0000	1.7800	0.1500
	Fei_vv_task1_4	Fei2023	23	28.25	51.8	7802348	70588	0.7595	8.0000	1.5800	0.1500
	Han_SZU_task1_1	Han2023	39	37.75	50.5	29.349M	80845	1.0845	308.0000	0.2840	0.0010
	LAM_AEV_task1_1	Pham2023	25	29.00	55.3	29267550	88704	0.0276	0.5152	82.7400	4.4270
	LAM_AEV_task1_2	Pham2023	31	31.50	55.0	29267550	88704	0.0138	0.5152	164.7400	4.4270
	LAM_AEV_task1_3	Pham2023	20	27.00	55.6	29267550	88704	0.0138	0.5152	164.7400	4.4270
	Liang_NTES_task1_1	Liang2023	40	38.50	52.6	29591778	143345		5.3234		0.1670
	MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	14686940	119608	0.0152	0.9375	2.9640	0.0480
	MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	10819292	87160	0.0185	0.9574	2.4310	0.0470
	MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	572340	116008	0.0507	5.7273	1.2420	0.0110
	MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	214420	63592	0.0098	1.1538	1.5300	0.0130
	DCASE2023 baseline		52	46.75	44.8	29234920	65280	0.9669	1.0000	0.3020	0.2920
	Park_KT_task1_1	Kim2023	10	20.75	56.3	19556096	250000	4.4282	56.5613	0.0530	0.0042
	Park_KT_task1_2	Kim2023	29	30.75	53.0	19556096	250000	4.4282	56.5613	0.0530	0.0042
	Park_KT_task1_3	Kim2023	26	29.25	54.2	19556096	250000	4.4282	56.5613	0.0530	0.0042
	Park_KT_task1_4	Kim2023	20	27.00	49.2	617000	30000	248.0005	1287.1071	0.0009	0.0002
	Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	1582336	5722	0.1249	7.1515	1.8890	0.0330
	Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	4354304	12310	0.1204	6.7429	1.9600	0.0350
	Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	9638144	30106	0.1111	6.3784	2.1250	0.0370
	Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	16803072	54182	0.0663	6.5556	3.5600	0.0360
	Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	28931380	127988	0.0678	50.1029	5.2587	0.0071
	Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	19910080	68456	0.0747	49.8043	4.7693	0.0072
	Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	9996775	74700	0.0745	44.6759	4.7830	0.0080
	Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	4938255	34616	0.0726	53.7101	4.9099	0.0066
	Tan_NTU_task1_1	Tan2023	33	33.50	47.1	2960384	40960	1.7114	8.8426	0.2692	0.0521
	Tan_NTU_task1_2	Tan2023	35	34.00	48.5	6462656	54304	1.0784	4.3258	0.4272	0.1065
	Tan_NTU_task1_3	Tan2023	37	37.00	46.3	6462656	54304	0.8352	4.3340	0.5516	0.1063
	Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	13180000	73386	0.0543	7.6842	5.3790	0.0380
	Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	13180000	73386	0.1614	9.6689	1.8090	0.0302
	Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	13180000	73386	0.0553	8.5882	5.2810	0.0340
	Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	13180000	73386	0.1614	9.1250	1.8090	0.0320
	Vo_DU_task1_1	Vo2023	52	46.75	45.0	15600000	503316	0.0064	0.6503	3.3220	0.0326
	Vo_DU_task1_2	Vo2023	53	47.75	44.8	15600000	503316	0.0068	0.9021	3.1220	0.0235
	Vo_DU_task1_3	Vo2023	50	46.25	45.2	15600000	503316	0.0071	1.8120	2.9780	0.0117
	Vo_DU_task1_4	Vo2023	49	45.75	45.5	15600000	503316	0.0121	17.6667	1.7560	0.0012
	Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	8646000	62080	1.0735	1.1542	0.2720	0.2530
	Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	16746000	81150	0.7449	0.7871	0.3920	0.3710
	Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	25442000	82280	0.8044	0.8439	0.3630	0.3460
	Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	20902000	148618	0.6173	0.8957	0.4730	0.3260
	XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	23803968	125885	0.0622	2.0603	1.3576	0.0410
	XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	28400320	123654	0.0504	2.0093	1.6751	0.0420
	XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	13402688	125650	0.0439	2.2005	1.9233	0.0384
	XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	11878580	125057	0.0439	2.2438	1.9208	0.0376
	Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	23970000	76906	0.0615	31.4878	21.0000	0.0410
	Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	23970000	76906	0.0615	5.3792	21.0000	0.2400
	Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	23970000	76906	0.0759	31.4878	17.0000	0.0410
	Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	23970000	76906	0.0759	5.3792	17.0000	0.2400
	Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	7375000	574464	0.8202	18.8125	0.3670	0.0160
	Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	28461000	1622016	0.7679	6.5435	0.3920	0.0460
	Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	3972096	7946	1.8354	5.8000	0.0790	0.0250
	Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	19466240	46232	0.2757	0.5142	0.5260	0.2820
	Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	23438336	54178	0.9477	3.1522	0.1530	0.0460
	Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	7944192	15892	0.2397	0.4723	0.6050	0.3070

Generalization performance

All results with evaluation dataset.

	Submission information		Overall				Devices				Cities
			Evaluation dataset				Unseen		Seen		Unseen		Seen
Rank	Submission label	Technical Report	Official system rank	Rank value	Accuracy (Evaluation dataset)	Logloss (Evaluation dataset)	Accuracy / unseen devices (Evaluation dataset)	Logloss / unseen devices (Evaluation dataset)	Accuracy / seen devices (Evaluation dataset)	Logloss / seen devices (Evaluation dataset)	Accuracy / unseen cities (Evaluation dataset)	Logloss / unseen cities (Evaluation dataset)	Accuracy / seen cities (Evaluation dataset)	Logloss / seen cities (Evaluation dataset)
	AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	1.920	47.0	2.370	56.1	1.544	51.6	1.931	52.3	1.905
	AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	1.996	42.4	2.391	54.1	1.667	49.2	1.847	48.8	2.005
	AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	1.364	45.3	1.528	55.4	1.227	52.5	1.302	50.7	1.372
	Bai_JLESS_task1_1	Du2023	54	50.25	47.9	1.825	39.5	2.335	55.0	1.399	44.2	1.944	48.9	1.799
	Bai_JLESS_task1_2	Du2023	47	43.75	48.8	1.791	40.6	2.225	55.6	1.430	44.0	1.817	50.0	1.794
	Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	1.174	50.7	1.347	61.5	1.029	54.9	1.202	57.1	1.167
	Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	1.246	51.8	1.369	59.8	1.143	55.1	1.275	56.5	1.242
	Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	1.241	51.9	1.356	59.1	1.144	55.3	1.255	56.0	1.241
	Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	1.252	51.3	1.358	58.7	1.164	55.7	1.259	55.3	1.255
	Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	1.307	49.1	1.390	54.2	1.237	50.9	1.325	52.1	1.306
	Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	1.292	49.7	1.389	55.0	1.211	53.5	1.285	52.5	1.298
	Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	1.223	51.6	1.331	58.0	1.133	56.2	1.218	55.0	1.226
	Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	1.241	53.5	1.385	59.9	1.121	56.8	1.259	57.5	1.234
	Fei_vv_task1_1	Fei2023	15	24.75	55.2	1.282	51.7	1.393	58.2	1.190	54.0	1.298	55.8	1.274
	Fei_vv_task1_2	Fei2023	13	21.75	54.5	1.290	50.9	1.410	57.5	1.191	52.9	1.296	55.2	1.281
	Fei_vv_task1_3	Fei2023	26	29.25	53.2	1.349	49.9	1.457	56.1	1.259	51.5	1.373	54.1	1.337
	Fei_vv_task1_4	Fei2023	23	28.25	51.8	1.370	48.8	1.460	54.2	1.295	49.9	1.389	52.6	1.359
	Han_SZU_task1_1	Han2023	39	37.75	50.5	2.012	44.9	2.565	55.2	1.550	49.2	1.966	51.1	2.020
	LAM_AEV_task1_1	Pham2023	25	29.00	55.3	1.847	49.6	2.329	60.1	1.445	55.4	1.884	55.6	1.849
	LAM_AEV_task1_2	Pham2023	31	31.50	55.0	2.083	48.9	2.681	60.1	1.584	54.1	2.134	55.5	2.074
	LAM_AEV_task1_3	Pham2023	20	27.00	55.6	1.933	49.6	2.447	60.5	1.504	54.9	1.937	56.0	1.938
	Liang_NTES_task1_1	Liang2023	40	38.50	52.6	1.402	44.9	1.704	58.9	1.150	51.3	1.408	53.0	1.397
	MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	1.230	52.3	1.349	60.9	1.130	56.1	1.239	57.4	1.227
	MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	1.242	51.9	1.353	60.4	1.150	55.5	1.265	57.1	1.235
	MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	4.354	10.0	4.358	9.8	4.350	9.9	4.360	9.9	4.349
	MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	3.224	9.8	3.228	9.9	3.221	10.0	3.225	9.8	3.223
	DCASE2023 baseline		52	46.75	44.8	1.523	38.0	1.730	50.5	1.351	42.4	1.563	45.6	1.513
	Park_KT_task1_1	Kim2023	10	20.75	56.3	1.495	49.9	1.827	61.7	1.218	54.4	1.543	57.1	1.471
	Park_KT_task1_2	Kim2023	29	30.75	53.0	1.660	46.2	2.056	58.7	1.329	51.7	1.720	53.9	1.629
	Park_KT_task1_3	Kim2023	26	29.25	54.2	2.230	47.1	2.840	60.2	1.721	53.6	2.240	54.7	2.212
	Park_KT_task1_4	Kim2023	20	27.00	49.2	1.510	45.8	1.573	52.0	1.457	49.2	1.480	49.4	1.512
	Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	1.313	52.3	1.395	56.1	1.244	53.7	1.318	54.5	1.317
	Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	1.256	54.6	1.445	62.1	1.099	59.4	1.215	58.8	1.264
	Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	1.153	57.8	1.318	64.4	1.015	60.8	1.133	61.8	1.156
	Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	1.117	57.4	1.334	67.1	0.936	61.2	1.152	63.4	1.107
	Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	1.322	51.0	1.465	59.6	1.203	55.3	1.336	56.0	1.322
	Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	1.337	50.8	1.488	59.5	1.210	55.6	1.347	55.8	1.336
	Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	1.398	47.6	1.562	57.0	1.261	52.3	1.408	52.9	1.396
	Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	1.482	44.8	1.650	53.8	1.342	50.0	1.489	50.1	1.477
	Tan_NTU_task1_1	Tan2023	33	33.50	47.1	1.508	39.7	1.720	53.3	1.331	45.8	1.509	47.6	1.503
	Tan_NTU_task1_2	Tan2023	35	34.00	48.5	1.461	41.4	1.634	54.4	1.317	47.4	1.466	49.2	1.454
	Tan_NTU_task1_3	Tan2023	37	37.00	46.3	1.492	40.1	1.632	51.5	1.375	44.2	1.517	47.1	1.482
	Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	1.192	55.7	1.420	65.1	1.002	59.3	1.249	61.5	1.180
	Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	1.444	45.8	1.726	56.6	1.208	50.9	1.481	52.1	1.434
	Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	1.441	48.6	1.675	57.5	1.247	52.9	1.487	53.9	1.431
	Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	1.525	45.5	1.843	55.4	1.259	51.1	1.542	51.1	1.519
	Vo_DU_task1_1	Vo2023	52	46.75	45.0	2.157	36.2	2.963	52.3	1.485	42.6	2.334	45.8	2.111
	Vo_DU_task1_2	Vo2023	53	47.75	44.8	2.116	36.3	2.877	51.9	1.482	42.3	2.289	45.7	2.069
	Vo_DU_task1_3	Vo2023	50	46.25	45.2	2.092	36.6	2.827	52.4	1.479	43.0	2.286	46.2	2.056
	Vo_DU_task1_4	Vo2023	49	45.75	45.5	1.793	38.0	2.169	51.8	1.479	43.7	1.793	46.3	1.771
	Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	1.493	41.8	1.823	55.3	1.218	47.5	1.530	49.8	1.478
	Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	1.348	47.3	1.510	57.5	1.214	52.0	1.375	53.4	1.339
	Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	1.702	38.4	2.228	54.4	1.263	45.2	1.830	47.7	1.680
	Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	1.472	42.9	1.654	53.1	1.321	45.1	1.556	49.4	1.455
	XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	1.364	47.5	1.450	53.9	1.292	50.3	1.377	51.0	1.365
	XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	1.355	47.4	1.455	55.1	1.271	51.0	1.369	51.6	1.357
	XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	1.395	45.5	1.509	53.8	1.301	49.5	1.408	50.2	1.396
	XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	1.367	47.6	1.450	54.1	1.298	49.4	1.400	51.6	1.363
	Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	1.280	52.6	1.409	58.0	1.172	55.6	1.278	55.9	1.276
	Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	1.241	53.2	1.352	58.2	1.149	56.0	1.242	56.3	1.237
	Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	1.279	52.1	1.406	57.6	1.172	56.7	1.253	55.2	1.283
	Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	1.259	52.4	1.364	57.5	1.172	56.8	1.226	55.2	1.263
	Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	1.757	33.3	2.083	51.6	1.485	40.8	1.831	44.0	1.736
	Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	1.533	39.7	1.813	54.7	1.300	46.4	1.570	48.2	1.530
	Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	3.248	39.2	5.085	56.8	1.717	48.3	3.481	49.2	3.229
	Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	4.213	40.6	6.653	57.7	2.179	48.7	4.433	50.4	4.145
	Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	1.704	41.8	2.237	60.3	1.261	50.7	1.823	52.4	1.679
	Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	1.542	41.6	1.802	57.6	1.325	49.8	1.585	50.7	1.532

Class-wise performance

Accuracy

Submission label	Technical Report	Official system rank	Rank value	Accuracy	Airport	Bus	Metro	Metro station	Park	Public square	Shopping mall	Street pedestrian	Street traffic	Tram
AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	45.9	72.9	46.4	44.6	73.2	26.7	59.7	27.9	57.3	64.7
AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	40.8	72.2	51.1	40.0	73.1	21.3	43.6	31.8	59.7	54.5
AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	49.4	67.4	52.5	38.1	82.0	28.4	39.2	31.3	56.5	63.2
Bai_JLESS_task1_1	Du2023	54	50.25	47.9	25.5	60.8	51.2	51.7	69.0	22.9	52.3	24.9	65.8	55.0
Bai_JLESS_task1_2	Du2023	47	43.75	48.8	29.4	67.1	49.8	44.5	70.1	25.9	51.9	22.9	70.2	56.2
Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	40.5	59.7	61.8	57.8	79.9	35.6	54.8	35.3	73.4	67.4
Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	34.4	64.5	58.6	62.0	83.4	36.0	56.7	31.1	68.8	66.3
Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	34.4	64.9	54.2	56.9	81.7	33.6	57.1	36.3	69.9	69.3
Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	34.6	60.5	55.1	58.6	78.9	35.5	54.0	36.1	71.9	68.5
Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	34.8	66.1	46.4	43.4	79.2	28.4	61.7	33.8	67.4	57.7
Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	31.6	66.0	50.0	48.1	78.7	31.1	61.4	33.6	66.4	58.5
Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	39.7	64.2	56.0	49.4	80.2	33.7	64.2	34.0	64.7	64.7
Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	44.7	68.7	57.6	52.5	83.0	36.8	52.4	34.5	70.5	69.4
Fei_vv_task1_1	Fei2023	15	24.75	55.2	45.6	71.9	58.1	45.8	76.3	31.0	59.7	31.7	67.0	65.3
Fei_vv_task1_2	Fei2023	13	21.75	54.5	37.1	73.4	54.7	45.1	77.7	28.1	60.4	36.6	67.5	64.4
Fei_vv_task1_3	Fei2023	26	29.25	53.2	57.9	84.5	54.4	42.5	73.1	24.9	54.2	27.8	62.0	51.3
Fei_vv_task1_4	Fei2023	23	28.25	51.8	56.6	83.2	54.8	38.8	79.2	21.2	59.5	18.2	61.7	44.4
Han_SZU_task1_1	Han2023	39	37.75	50.5	39.7	60.8	54.6	43.5	80.6	28.0	51.6	32.1	62.8	51.8
LAM_AEV_task1_1	Pham2023	25	29.00	55.3	50.1	80.2	52.2	40.9	74.9	43.8	57.8	23.6	69.4	60.4
LAM_AEV_task1_2	Pham2023	31	31.50	55.0	35.6	79.9	44.5	54.3	82.3	33.4	58.0	32.0	68.4	61.7
LAM_AEV_task1_3	Pham2023	20	27.00	55.6	47.7	77.1	48.6	52.8	78.1	38.0	56.8	27.0	67.0	62.9
Liang_NTES_task1_1	Liang2023	40	38.50	52.6	38.8	72.7	47.9	45.1	74.8	27.2	56.8	36.5	56.2	69.7
MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	34.8	72.3	60.5	55.4	85.2	37.0	51.0	35.3	68.9	69.6
MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	34.8	69.7	54.0	57.8	81.4	37.0	51.2	40.2	69.3	70.0
MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	12.0	11.7	8.7	9.4	12.4	8.2	11.6	12.3	7.4	5.2
MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	16.6	10.3	7.2	9.3	11.9	6.5	13.0	4.8	9.8	9.1
DCASE2023 baseline		52	46.75	44.8	40.5	37.8	48.8	38.9	58.2	22.4	54.1	29.8	58.0	59.7
Park_KT_task1_1	Kim2023	10	20.75	56.3	36.2	80.8	59.8	55.1	70.1	29.6	52.6	37.4	69.5	72.1
Park_KT_task1_2	Kim2023	29	30.75	53.0	26.6	75.3	53.1	49.7	66.5	29.7	62.2	35.1	63.8	68.3
Park_KT_task1_3	Kim2023	26	29.25	54.2	38.1	78.5	57.0	52.5	73.1	29.4	45.4	33.3	67.9	67.1
Park_KT_task1_4	Kim2023	20	27.00	49.2	48.7	75.7	48.3	28.2	76.2	29.5	46.4	30.3	49.9	59.0
Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	56.4	72.1	58.0	44.2	82.9	32.1	58.7	14.7	70.2	54.3
Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	49.9	74.0	59.4	48.6	87.2	35.7	59.9	26.4	70.7	75.4
Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	50.4	78.0	65.6	60.3	84.7	35.1	65.7	33.6	69.6	71.3
Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	48.0	88.1	70.1	62.8	84.3	35.7	56.9	35.7	75.6	69.8
Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	40.0	66.2	59.3	53.7	78.8	36.8	56.3	31.7	71.9	62.2
Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	32.5	65.4	61.1	55.1	86.9	32.3	47.7	41.8	69.0	64.1
Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	34.0	67.6	59.7	51.8	78.1	25.4	53.3	37.1	64.5	55.2
Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	31.5	59.8	61.0	49.2	68.5	31.2	67.8	17.3	57.3	53.5
Tan_NTU_task1_1	Tan2023	33	33.50	47.1	33.8	63.3	41.0	39.7	62.8	21.8	58.3	28.5	61.4	60.3
Tan_NTU_task1_2	Tan2023	35	34.00	48.5	35.8	71.5	44.0	37.1	70.5	23.1	60.7	25.7	61.7	54.7
Tan_NTU_task1_3	Tan2023	37	37.00	46.3	40.6	55.1	45.2	48.2	75.9	13.7	58.4	14.1	63.6	48.5
Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	42.7	76.8	62.5	49.7	83.6	44.3	60.1	43.4	73.5	71.7
Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	37.7	70.8	47.8	42.0	75.3	32.7	50.0	32.5	67.0	61.3
Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	38.5	70.9	49.3	39.8	75.0	40.0	56.4	34.0	68.4	62.5
Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	35.5	79.1	42.7	49.4	74.3	31.3	42.9	38.1	65.7	50.2
Vo_DU_task1_1	Vo2023	52	46.75	45.0	45.6	58.1	40.8	39.1	65.9	20.0	39.6	26.7	62.5	51.3
Vo_DU_task1_2	Vo2023	53	47.75	44.8	42.6	59.5	40.9	40.2	64.9	22.0	35.8	28.1	65.2	48.8
Vo_DU_task1_3	Vo2023	50	46.25	45.2	45.2	60.1	38.6	42.3	68.4	21.9	40.9	23.8	61.8	49.2
Vo_DU_task1_4	Vo2023	49	45.75	45.5	41.5	50.4	57.0	29.1	71.7	16.6	43.8	34.0	61.7	49.5
Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	52.9	68.7	43.2	29.2	71.3	23.1	42.3	34.5	58.5	67.7
Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	58.1	63.7	44.0	28.7	71.7	35.4	54.1	34.1	63.1	75.8
Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	24.5	66.6	62.7	69.1	68.4	21.1	36.7	24.4	63.3	34.5
Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	45.9	48.3	37.3	49.7	79.3	23.5	55.1	23.2	65.8	56.6
XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	16.2	59.7	52.8	54.0	81.3	30.4	55.8	42.3	65.2	52.4
XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	22.0	73.9	47.4	57.2	79.9	33.5	53.5	33.9	60.7	53.8
XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	22.4	67.5	50.1	56.4	82.3	31.6	52.1	26.4	59.0	52.6
XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	24.2	67.2	46.5	59.4	75.2	31.8	58.2	33.0	64.5	51.5
Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	40.3	62.4	60.1	50.2	77.2	33.2	60.6	32.8	73.9	64.7
Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	40.8	60.9	58.9	48.9	78.9	34.7	61.2	32.6	74.1	68.2
Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	42.1	74.0	52.2	43.3	82.7	36.9	62.2	26.5	68.9	62.4
Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	44.7	72.8	49.4	42.4	82.2	35.3	61.5	28.0	70.4	65.1
Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	26.0	54.3	49.7	43.6	62.7	18.8	50.2	17.1	60.8	49.6
Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	26.9	71.4	44.0	35.1	79.9	32.3	46.5	33.9	59.1	49.7
Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	35.6	76.1	39.6	36.0	77.7	33.4	50.4	34.5	57.4	47.1
Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	37.5	72.7	46.6	39.9	76.1	29.9	51.9	31.8	58.6	54.7
Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	36.7	78.5	48.9	39.8	82.0	31.3	54.1	38.9	54.9	53.9
Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	33.6	75.0	42.7	37.4	84.9	29.4	52.1	41.7	55.3	51.0

Log loss

Submission label	Technical Report	Official system rank	Rank value	Logloss	Airport	Bus	Metro	Metro station	Park	Public square	Shopping mall	Street pedestrian	Street traffic	Tram
AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	1.920	1.521	1.162	1.712	2.025	2.107	2.513	2.220	2.167	2.708	1.062
AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	1.996	1.979	1.132	1.686	1.710	1.728	2.664	3.129	2.401	2.236	1.295
AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	1.364	1.382	0.939	1.269	1.773	0.632	1.946	1.623	1.773	1.293	1.007
Bai_JLESS_task1_1	Du2023	54	50.25	1.825	2.324	1.215	1.402	1.546	1.724	2.876	1.531	2.857	1.436	1.336
Bai_JLESS_task1_2	Du2023	47	43.75	1.791	2.241	1.049	1.574	1.965	1.239	2.638	1.684	2.946	1.194	1.379
Cai_TENCENT_task1_1	Cai2023	17	25.75	1.174	1.446	1.060	0.971	1.180	0.646	1.776	1.219	1.695	0.862	0.883
Cai_TENCENT_task1_2	Cai2023	11	21.25	1.246	1.628	0.988	1.103	1.149	0.629	1.805	1.266	1.903	1.039	0.948
Cai_TENCENT_task1_3	Cai2023	14	22.25	1.241	1.598	0.986	1.188	1.275	0.666	1.775	1.240	1.770	1.013	0.894
Cai_TENCENT_task1_4	Cai2023	11	21.25	1.252	1.599	1.096	1.161	1.244	0.766	1.723	1.307	1.752	0.954	0.918
Cai_XJTLU_task1_1	Cai2023a	9	19.75	1.307	1.535	1.008	1.317	1.596	0.750	1.859	1.153	1.715	1.005	1.129
Cai_XJTLU_task1_2	Cai2023a	8	18.25	1.292	1.609	0.974	1.221	1.451	0.750	1.826	1.150	1.761	1.077	1.096
Cai_XJTLU_task1_3	Cai2023a	6	14.00	1.223	1.446	0.998	1.103	1.394	0.655	1.824	1.043	1.729	1.107	0.932
Cai_XJTLU_task1_4	Cai2023a	3	11.50	1.241	1.502	0.908	1.126	1.429	0.562	1.873	1.320	1.889	0.955	0.849
Fei_vv_task1_1	Fei2023	15	24.75	1.282	1.418	0.835	1.178	1.567	0.753	1.959	1.141	1.884	1.100	0.986
Fei_vv_task1_2	Fei2023	13	21.75	1.290	1.636	0.796	1.238	1.551	0.729	1.977	1.146	1.719	1.117	0.995
Fei_vv_task1_3	Fei2023	26	29.25	1.349	1.592	0.695	1.491	1.644	0.930	2.276	1.159	1.603	1.031	1.069
Fei_vv_task1_4	Fei2023	23	28.25	1.370	1.671	0.818	1.469	1.730	0.702	2.297	0.941	1.847	1.003	1.226
Han_SZU_task1_1	Han2023	39	37.75	2.012	2.062	1.571	1.586	2.090	0.780	3.228	2.380	2.773	2.029	1.616
LAM_AEV_task1_1	Pham2023	25	29.00	1.847	1.943	0.801	1.551	2.243	1.235	2.526	1.813	3.416	1.577	1.364
LAM_AEV_task1_2	Pham2023	31	31.50	2.083	2.826	0.828	2.244	1.996	0.842	3.829	2.003	3.330	1.515	1.416
LAM_AEV_task1_3	Pham2023	20	27.00	1.933	2.060	0.936	1.820	1.912	1.074	3.087	1.947	3.500	1.652	1.339
Liang_NTES_task1_1	Liang2023	40	38.50	1.402	1.610	0.783	1.515	1.710	0.845	2.014	1.272	1.822	1.588	0.861
MALACH23_JKU_task1_1	Pichler2023	8	18.25	1.230	1.676	0.883	1.110	1.313	0.596	1.672	1.381	1.755	0.997	0.913
MALACH23_JKU_task1_2	Pichler2023	7	16.75	1.242	1.661	0.953	1.239	1.262	0.716	1.637	1.350	1.659	1.000	0.945
MALACH23_JKU_task1_3	Pichler2023	42	39.50	4.354	4.368	4.889	4.315	3.618	5.426	3.395	4.872	3.433	4.845	4.376
MALACH23_JKU_task1_4	Pichler2023	43	39.75	3.224	3.160	3.690	3.201	2.570	3.634	2.692	3.373	2.840	3.927	3.157
DCASE2023 baseline		52	46.75	1.523	1.487	1.585	1.294	1.689	1.452	2.085	1.360	1.817	1.365	1.100
Park_KT_task1_1	Kim2023	10	20.75	1.495	2.157	0.591	1.287	1.530	1.067	2.541	1.623	2.210	1.106	0.836
Park_KT_task1_2	Kim2023	29	30.75	1.660	2.786	0.766	1.556	1.807	1.128	2.542	1.342	2.335	1.342	0.993
Park_KT_task1_3	Kim2023	26	29.25	2.230	2.871	0.933	2.170	2.357	1.302	3.469	2.722	3.498	1.467	1.509
Park_KT_task1_4	Kim2023	20	27.00	1.510	1.618	1.014	1.459	1.883	1.047	1.889	1.598	1.797	1.535	1.254
Schmid_CPJKU_task1_1	Schmid2023	5	13.75	1.313	1.219	0.901	1.115	1.634	0.584	2.156	1.080	2.250	0.989	1.198
Schmid_CPJKU_task1_2	Schmid2023	1	5.25	1.256	1.421	0.842	1.175	1.596	0.425	2.134	1.140	2.266	0.906	0.657
Schmid_CPJKU_task1_3	Schmid2023	2	7.00	1.153	1.364	0.697	1.009	1.193	0.470	2.095	0.960	1.970	0.988	0.783
Schmid_CPJKU_task1_4	Schmid2023	3	11.50	1.117	1.498	0.388	0.873	1.087	0.518	2.034	1.233	1.950	0.774	0.814
Schmidt_FAU_task1_1	Schmidt2023	26	29.25	1.322	1.770	1.142	1.253	1.397	0.802	1.839	1.263	1.801	0.949	1.007
Schmidt_FAU_task1_2	Schmidt2023	12	21.50	1.337	1.930	1.175	1.199	1.374	0.583	1.982	1.532	1.568	1.041	0.980
Schmidt_FAU_task1_3	Schmidt2023	19	26.75	1.398	1.837	1.101	1.207	1.449	0.845	2.227	1.311	1.671	1.191	1.137
Schmidt_FAU_task1_4	Schmidt2023	24	28.75	1.482	2.030	1.351	1.158	1.477	1.132	2.101	1.012	2.060	1.370	1.127
Tan_NTU_task1_1	Tan2023	33	33.50	1.508	1.675	1.089	1.389	1.687	1.253	2.095	1.402	2.048	1.316	1.124
Tan_NTU_task1_2	Tan2023	35	34.00	1.461	1.614	0.956	1.376	1.719	1.078	2.011	1.408	1.973	1.282	1.192
Tan_NTU_task1_3	Tan2023	37	37.00	1.492	1.547	1.311	1.363	1.519	1.044	2.122	1.392	2.074	1.181	1.363
Tan_SCUT_task1_1	Tan2023a	4	13.50	1.192	1.627	0.664	1.014	1.608	0.537	1.663	1.148	2.026	0.818	0.818
Tan_SCUT_task1_2	Tan2023a	30	31.00	1.444	1.789	0.889	1.294	1.656	0.864	1.979	1.537	2.237	1.116	1.075
Tan_SCUT_task1_3	Tan2023a	16	25.50	1.441	1.767	0.869	1.312	1.922	0.912	1.908	1.320	2.234	1.055	1.112
Tan_SCUT_task1_4	Tan2023a	32	33.00	1.525	1.938	0.654	1.470	1.528	0.993	2.129	1.728	2.201	1.202	1.405
Vo_DU_task1_1	Vo2023	52	46.75	2.157	1.492	1.532	2.737	1.898	1.042	3.724	2.305	4.132	1.336	1.373
Vo_DU_task1_2	Vo2023	53	47.75	2.116	1.603	1.474	2.640	1.871	1.086	3.194	2.548	4.103	1.230	1.412
Vo_DU_task1_3	Vo2023	50	46.25	2.092	1.541	1.508	2.746	1.762	0.993	3.112	2.224	4.321	1.325	1.387
Vo_DU_task1_4	Vo2023	49	45.75	1.793	1.633	1.567	1.534	2.211	0.976	3.063	1.866	2.395	1.295	1.390
Wang_SCUT_task1_1	Wang2023	31	31.50	1.493	1.234	0.904	1.480	2.014	1.130	2.168	1.615	2.120	1.312	0.957
Wang_SCUT_task1_2	Wang2023	18	26.50	1.348	1.130	0.989	1.427	2.059	1.186	1.579	1.191	1.818	1.317	0.788
Wang_SCUT_task1_3	Wang2023	46	43.50	1.702	2.422	0.984	1.010	0.940	1.218	2.515	2.111	2.914	1.289	1.614
Wang_SCUT_task1_4	Wang2023	48	45.00	1.472	1.363	1.565	1.891	1.454	0.760	1.956	1.241	2.119	1.053	1.321
XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	1.364	1.867	1.240	1.337	1.412	0.716	1.842	1.278	1.569	1.140	1.235
XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	1.355	1.827	0.902	1.344	1.339	0.746	1.851	1.326	1.764	1.279	1.170
XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	1.395	1.845	1.033	1.291	1.324	0.678	1.955	1.348	1.963	1.314	1.202
XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	1.367	1.743	1.083	1.413	1.299	0.875	1.877	1.214	1.738	1.177	1.255
Yang_GZHU_task1_1	Weng2023	15	24.75	1.280	1.519	1.091	1.098	1.464	0.741	2.027	1.113	1.916	0.861	0.968
Yang_GZHU_task1_2	Weng2023	14	22.25	1.241	1.467	1.099	1.100	1.475	0.680	1.920	1.111	1.867	0.845	0.849
Yang_GZHU_task1_3	Weng2023	21	27.25	1.279	1.443	0.753	1.287	1.630	0.558	1.981	1.092	2.067	0.970	1.005
Yang_GZHU_task1_4	Weng2023	19	26.75	1.259	1.378	0.795	1.340	1.627	0.591	1.937	1.120	1.972	0.913	0.920
Zhang_NCUT_task1_1	Zhang2023	49	45.75	1.757	2.159	1.382	1.482	1.701	1.165	2.575	1.538	2.788	1.292	1.484
Zhang_NCUT_task1_2	Zhang2023	51	46.50	1.533	2.046	0.927	1.587	1.903	0.735	2.025	1.641	1.830	1.301	1.337
Zhang_SATLab_task1_1	Bing2023	26	29.25	3.248	4.626	1.072	2.657	4.508	1.552	3.713	3.538	5.323	3.042	2.446
Zhang_SATLab_task1_2	Bing2023	34	33.75	4.213	4.392	1.559	3.338	4.560	2.582	6.580	4.340	7.767	4.126	2.881
Zhang_SATLab_task1_3	Bing2023	27	30.00	1.704	1.810	0.713	1.606	2.119	0.817	2.119	1.555	1.766	2.093	2.443
Zhang_SATLab_task1_4	Bing2023	22	27.50	1.542	1.876	0.816	1.767	1.824	0.643	2.127	1.511	1.712	1.551	1.595

Device-wise performance

Accuracy

								Unseen devices					Seen devices
Rank	Submission label	Technical Report	Official system rank	Rank value	Accuracy	Accuracy / Unseen	Accuracy / Seen	D	S7	S8	S9	S10	A	B	C	S1	S2	S3
	AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	47.0	56.1	32.2	51.5	50.8	50.5	49.8	59.3	55.9	58.6	56.5	51.1	55.1
	AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	42.4	54.1	30.3	46.6	44.9	44.1	46.1	56.0	55.1	56.2	52.8	50.8	53.9
	AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	45.3	55.4	42.7	46.4	45.9	44.6	46.9	58.2	55.2	56.5	54.4	52.2	55.9
	Bai_JLESS_task1_1	Du2023	54	50.25	47.9	39.5	55.0	29.8	48.8	47.3	34.3	37.3	64.5	56.0	60.6	49.3	49.3	50.0
	Bai_JLESS_task1_2	Du2023	47	43.75	48.8	40.6	55.6	31.5	48.6	47.8	35.1	40.0	64.6	56.6	61.8	49.9	49.8	50.8
	Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	50.7	61.5	46.0	57.2	58.0	45.1	47.4	69.1	60.3	65.0	57.7	57.9	59.2
	Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	51.8	59.8	37.6	57.7	57.6	53.4	52.9	66.0	57.6	61.5	57.0	58.3	58.4
	Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	51.9	59.1	39.3	58.0	57.1	52.5	52.5	65.8	56.3	61.7	56.3	56.3	58.4
	Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	51.3	58.7	39.2	56.5	55.8	52.8	52.4	65.1	56.7	61.2	56.1	56.0	57.3
	Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	49.1	54.2	44.6	54.3	52.7	46.5	47.3	60.6	52.7	56.2	52.7	50.4	52.7
	Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	49.7	55.0	44.0	54.5	54.5	47.6	47.7	62.6	51.6	56.5	53.6	51.3	54.1
	Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	51.6	58.0	42.9	57.3	57.2	50.4	50.3	65.0	54.7	58.7	57.2	55.6	56.7
	Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	53.5	59.9	46.2	58.3	58.4	52.7	51.9	66.5	57.5	59.8	57.6	59.0	59.2
	Fei_vv_task1_1	Fei2023	15	24.75	55.2	51.7	58.2	44.9	57.6	53.7	53.1	49.4	64.8	57.1	60.4	56.0	55.1	55.8
	Fei_vv_task1_2	Fei2023	13	21.75	54.5	50.9	57.5	42.8	56.8	54.0	52.3	48.4	64.8	55.1	59.4	56.7	53.9	55.2
	Fei_vv_task1_3	Fei2023	26	29.25	53.2	49.9	56.1	42.6	55.6	53.3	50.8	47.0	62.2	55.9	58.8	54.1	53.0	52.4
	Fei_vv_task1_4	Fei2023	23	28.25	51.8	48.8	54.2	41.2	54.2	53.7	48.8	46.3	60.5	52.1	56.8	53.5	50.7	51.5
	Han_SZU_task1_1	Han2023	39	37.75	50.5	44.9	55.2	41.0	51.3	50.0	38.3	43.8	65.0	52.6	59.0	51.9	50.0	53.1
	LAM_AEV_task1_1	Pham2023	25	29.00	55.3	49.6	60.1	43.6	57.7	57.3	39.4	50.3	67.8	56.6	64.3	58.7	53.8	59.3
	LAM_AEV_task1_2	Pham2023	31	31.50	55.0	48.9	60.1	44.7	56.0	54.5	38.8	50.6	68.0	56.4	63.8	56.9	56.4	58.9
	LAM_AEV_task1_3	Pham2023	20	27.00	55.6	49.6	60.5	44.6	58.7	55.4	39.7	49.8	69.2	56.4	64.6	57.9	56.0	59.1
	Liang_NTES_task1_1	Liang2023	40	38.50	52.6	44.9	58.9	32.5	54.4	53.9	39.8	44.0	67.5	58.2	63.7	57.0	52.2	55.0
	MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	52.3	60.9	44.7	57.7	59.1	49.5	50.6	68.4	57.3	63.5	58.0	57.9	60.2
	MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	51.9	60.4	42.9	56.6	58.3	51.4	50.4	67.7	56.7	63.1	57.7	57.7	59.7
	MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	10.0	9.8	9.8	10.1	9.6	9.9	10.6	10.1	9.9	9.8	9.9	9.5	9.8
	MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	9.8	9.9	9.3	10.2	9.9	9.6	10.0	9.9	9.9	9.8	9.8	9.6	10.2
	DCASE2023 baseline		52	46.75	44.8	38.0	50.5	34.7	46.2	43.6	31.1	34.2	60.4	50.1	54.9	45.1	45.0	47.7
	Park_KT_task1_1	Kim2023	10	20.75	56.3	49.9	61.7	33.3	60.3	59.3	50.3	46.4	68.8	60.7	63.8	59.0	57.6	60.2
	Park_KT_task1_2	Kim2023	29	30.75	53.0	46.2	58.7	34.6	54.9	52.5	45.7	43.2	64.1	58.4	61.3	56.2	56.2	56.2
	Park_KT_task1_3	Kim2023	26	29.25	54.2	47.1	60.2	32.6	56.6	56.2	47.0	43.2	67.3	59.0	61.3	57.3	56.8	59.3
	Park_KT_task1_4	Kim2023	20	27.00	49.2	45.8	52.0	41.9	50.0	48.8	44.2	44.3	60.4	48.8	52.7	51.9	46.8	51.5
	Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	52.3	56.1	42.9	56.8	55.9	53.1	53.0	59.7	53.7	57.5	56.0	53.8	55.5
	Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	54.6	62.1	47.9	59.6	58.5	52.4	54.6	66.9	59.5	63.1	61.0	60.7	61.6
	Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	57.8	64.4	48.1	63.3	62.1	58.7	57.0	68.8	61.2	64.5	64.7	63.3	64.0
	Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	57.4	67.1	45.1	63.9	64.1	56.4	57.6	71.0	64.9	67.9	66.4	66.6	66.0
	Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	51.0	59.6	40.5	57.5	56.2	51.0	49.6	66.5	57.6	63.1	56.6	56.7	57.3
	Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	50.8	59.5	38.4	55.5	56.3	52.2	51.7	66.8	56.2	62.3	56.9	56.5	58.4
	Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	47.6	57.0	37.7	55.2	54.2	44.4	46.3	64.5	54.5	61.0	53.0	53.6	55.0
	Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	44.8	53.8	38.1	52.2	51.6	38.7	43.6	61.7	52.2	55.5	53.1	48.7	51.3
	Tan_NTU_task1_1	Tan2023	33	33.50	47.1	39.7	53.3	36.5	49.3	40.5	35.8	36.4	63.3	51.4	57.7	48.4	47.6	51.2
	Tan_NTU_task1_2	Tan2023	35	34.00	48.5	41.4	54.4	36.0	50.1	42.2	37.3	41.4	64.2	50.6	58.8	50.0	49.1	53.5
	Tan_NTU_task1_3	Tan2023	37	37.00	46.3	40.1	51.5	29.5	47.0	45.8	37.8	40.4	61.0	45.5	55.9	47.7	48.3	50.6
	Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	55.7	65.1	42.8	62.6	60.0	56.3	57.1	70.0	61.5	66.2	64.1	64.1	64.6
	Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	45.8	56.6	33.5	55.4	54.2	41.3	44.8	62.9	52.5	57.9	55.0	55.1	56.1
	Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	48.6	57.5	36.2	56.9	55.6	47.8	46.5	62.8	53.8	58.7	56.3	55.6	58.0
	Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	45.5	55.4	30.9	54.6	52.8	42.2	46.9	59.4	51.7	57.1	53.9	54.4	56.1
	Vo_DU_task1_1	Vo2023	52	46.75	45.0	36.2	52.3	26.8	44.0	41.0	33.3	35.8	62.3	53.2	56.9	48.3	46.4	46.6
	Vo_DU_task1_2	Vo2023	53	47.75	44.8	36.3	51.9	28.4	43.5	40.0	33.8	35.5	62.3	52.8	57.1	47.4	46.1	45.8
	Vo_DU_task1_3	Vo2023	50	46.25	45.2	36.6	52.4	27.6	44.3	41.2	34.7	35.4	62.1	53.2	57.7	47.3	47.0	46.9
	Vo_DU_task1_4	Vo2023	49	45.75	45.5	38.0	51.8	29.5	43.2	44.7	36.5	35.9	61.7	51.3	56.9	46.3	46.4	48.3
	Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	41.8	55.3	26.1	51.3	48.0	39.3	44.1	61.5	52.6	57.5	54.8	52.9	52.6
	Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	47.3	57.5	38.8	55.9	53.5	42.1	46.3	61.8	54.8	60.1	57.0	56.2	55.3
	Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	38.4	54.4	22.8	49.1	48.1	39.3	32.9	61.9	53.1	56.8	48.6	53.5	52.2
	Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	42.9	53.1	33.1	50.5	50.1	38.0	42.9	59.8	50.2	53.4	53.2	51.6	50.5
	XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	47.5	53.9	41.4	51.3	52.3	46.4	46.2	58.0	53.6	57.2	52.5	50.1	52.1
	XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	47.4	55.1	43.5	50.2	52.7	45.0	45.4	58.4	55.7	58.8	53.5	51.4	52.7
	XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	45.5	53.8	41.5	48.8	50.0	42.0	45.2	58.5	53.6	57.7	51.3	49.6	52.0
	XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	47.6	54.1	43.4	50.5	50.6	47.2	46.2	59.4	55.3	58.0	51.6	48.9	51.4
	Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	52.6	58.0	46.9	58.4	55.6	51.8	50.3	64.1	56.4	60.6	55.6	55.6	55.7
	Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	53.2	58.2	47.2	58.7	56.4	52.5	51.2	63.5	56.8	61.0	56.2	55.6	55.8
	Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	52.1	57.6	42.2	57.3	56.2	54.7	50.0	62.5	55.7	61.4	56.5	54.7	55.0
	Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	52.4	57.5	43.4	57.5	56.1	54.6	50.6	62.3	56.2	61.1	56.1	54.3	54.7
	Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	33.3	51.6	34.0	38.8	38.0	27.6	27.8	63.2	55.0	57.6	45.1	45.9	43.0
	Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	39.7	54.7	37.9	48.3	48.0	27.0	37.4	63.0	54.8	61.0	49.4	50.0	50.0
	Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	39.2	56.8	32.2	46.7	37.9	27.7	51.4	64.0	54.1	61.1	54.0	53.6	53.8
	Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	40.6	57.7	28.6	50.5	47.5	30.0	46.7	64.7	57.5	62.9	54.3	53.3	53.8
	Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	41.8	60.3	27.2	51.8	46.9	32.7	50.6	67.5	59.4	65.2	57.0	56.4	56.4
	Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	41.6	57.6	34.0	47.2	43.7	33.3	49.6	65.0	55.7	62.4	54.0	54.0	54.5

Log loss

								Unseen devices					Seen devices
Rank	Submission label	Technical Report	Official system rank	Rank value	Log loss	Log loss / Unseen	Log loss / Seen	D	S7	S8	S9	S10	A	B	C	S1	S2	S3
	AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	1.920	2.370	1.544	4.428	1.659	1.969	1.831	1.965	1.325	1.601	1.421	1.480	1.815	1.622
	AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	1.996	2.391	1.667	3.613	1.860	2.071	2.198	2.211	1.523	1.672	1.468	1.729	1.848	1.762
	AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	1.364	1.528	1.227	1.568	1.485	1.534	1.565	1.490	1.156	1.215	1.183	1.261	1.319	1.225
	Bai_JLESS_task1_1	Du2023	54	50.25	1.825	2.335	1.399	3.859	1.607	1.631	2.131	2.448	1.095	1.349	1.163	1.579	1.650	1.560
	Bai_JLESS_task1_2	Du2023	47	43.75	1.791	2.225	1.430	2.609	1.820	1.623	2.561	2.510	1.129	1.303	1.160	1.654	1.669	1.662
	Cai_TENCENT_task1_1	Cai2023	17	25.75	1.174	1.347	1.029	1.463	1.168	1.138	1.513	1.454	0.850	1.071	0.929	1.113	1.123	1.088
	Cai_TENCENT_task1_2	Cai2023	11	21.25	1.246	1.369	1.143	1.843	1.221	1.198	1.258	1.326	1.003	1.191	1.107	1.214	1.177	1.166
	Cai_TENCENT_task1_3	Cai2023	14	22.25	1.241	1.356	1.144	1.768	1.195	1.191	1.296	1.332	0.996	1.200	1.094	1.206	1.208	1.159
	Cai_TENCENT_task1_4	Cai2023	11	21.25	1.252	1.358	1.164	1.695	1.230	1.218	1.301	1.345	1.019	1.200	1.112	1.234	1.219	1.200
	Cai_XJTLU_task1_1	Cai2023a	9	19.75	1.307	1.390	1.237	1.538	1.248	1.267	1.465	1.431	1.082	1.284	1.187	1.290	1.317	1.265
	Cai_XJTLU_task1_2	Cai2023a	8	18.25	1.292	1.389	1.211	1.605	1.247	1.231	1.432	1.430	1.029	1.275	1.158	1.263	1.304	1.235
	Cai_XJTLU_task1_3	Cai2023a	6	14.00	1.223	1.331	1.133	1.620	1.165	1.166	1.336	1.367	0.957	1.222	1.107	1.153	1.198	1.163
	Cai_XJTLU_task1_4	Cai2023a	3	11.50	1.241	1.385	1.121	1.779	1.192	1.200	1.358	1.398	0.953	1.193	1.130	1.169	1.140	1.144
	Fei_vv_task1_1	Fei2023	15	24.75	1.282	1.393	1.190	1.613	1.189	1.302	1.339	1.519	0.995	1.218	1.112	1.257	1.298	1.261
	Fei_vv_task1_2	Fei2023	13	21.75	1.290	1.410	1.191	1.650	1.212	1.304	1.361	1.522	0.990	1.246	1.128	1.232	1.313	1.236
	Fei_vv_task1_3	Fei2023	26	29.25	1.349	1.457	1.259	1.617	1.263	1.406	1.415	1.583	1.067	1.268	1.175	1.329	1.377	1.339
	Fei_vv_task1_4	Fei2023	23	28.25	1.370	1.460	1.295	1.614	1.297	1.419	1.418	1.555	1.083	1.355	1.236	1.321	1.432	1.346
	Han_SZU_task1_1	Han2023	39	37.75	2.012	2.565	1.550	2.860	1.664	1.923	4.114	2.263	1.370	1.748	1.420	1.574	1.721	1.469
	LAM_AEV_task1_1	Pham2023	25	29.00	1.847	2.329	1.445	3.322	1.521	1.618	3.158	2.026	1.282	1.655	1.289	1.463	1.567	1.416
	LAM_AEV_task1_2	Pham2023	31	31.50	2.083	2.681	1.584	3.367	1.903	2.028	3.894	2.215	1.350	1.745	1.425	1.688	1.729	1.566
	LAM_AEV_task1_3	Pham2023	20	27.00	1.933	2.447	1.504	3.117	1.577	1.816	3.558	2.166	1.218	1.781	1.304	1.570	1.631	1.521
	Liang_NTES_task1_1	Liang2023	40	38.50	1.402	1.704	1.150	2.196	1.328	1.364	1.938	1.694	0.922	1.154	0.995	1.207	1.357	1.266
	MALACH23_JKU_task1_1	Pichler2023	8	18.25	1.230	1.349	1.130	1.513	1.211	1.194	1.428	1.399	0.957	1.207	1.081	1.195	1.212	1.130
	MALACH23_JKU_task1_2	Pichler2023	7	16.75	1.242	1.353	1.150	1.558	1.236	1.205	1.365	1.402	0.974	1.226	1.099	1.220	1.218	1.161
	MALACH23_JKU_task1_3	Pichler2023	42	39.50	4.354	4.358	4.350	4.363	4.339	4.366	4.391	4.330	4.340	4.341	4.367	4.359	4.339	4.355
	MALACH23_JKU_task1_4	Pichler2023	43	39.75	3.224	3.228	3.221	3.245	3.203	3.226	3.249	3.216	3.215	3.224	3.214	3.225	3.225	3.225
	DCASE2023 baseline		52	46.75	1.523	1.730	1.351	1.847	1.476	1.535	1.929	1.862	1.101	1.366	1.262	1.448	1.500	1.430
	Park_KT_task1_1	Kim2023	10	20.75	1.495	1.827	1.218	2.779	1.275	1.410	1.746	1.927	1.024	1.210	1.103	1.310	1.356	1.302
	Park_KT_task1_2	Kim2023	29	30.75	1.660	2.056	1.329	2.983	1.509	1.699	1.941	2.150	1.190	1.315	1.198	1.454	1.396	1.423
	Park_KT_task1_3	Kim2023	26	29.25	2.230	2.840	1.721	4.363	2.054	2.076	2.688	3.021	1.436	1.762	1.608	1.816	1.888	1.816
	Park_KT_task1_4	Kim2023	20	27.00	1.510	1.573	1.457	1.614	1.496	1.530	1.642	1.581	1.276	1.526	1.422	1.488	1.561	1.468
	Schmid_CPJKU_task1_1	Schmid2023	5	13.75	1.313	1.395	1.244	1.822	1.216	1.244	1.341	1.353	1.120	1.360	1.191	1.226	1.311	1.255
	Schmid_CPJKU_task1_2	Schmid2023	1	5.25	1.256	1.445	1.099	1.833	1.188	1.267	1.502	1.437	0.962	1.202	1.075	1.128	1.123	1.101
	Schmid_CPJKU_task1_3	Schmid2023	2	7.00	1.153	1.318	1.015	1.840	1.056	1.153	1.252	1.290	0.892	1.134	1.000	1.018	1.038	1.010
	Schmid_CPJKU_task1_4	Schmid2023	3	11.50	1.117	1.334	0.936	1.902	1.044	1.045	1.310	1.366	0.847	1.001	0.919	0.949	0.935	0.966
	Schmidt_FAU_task1_1	Schmidt2023	26	29.25	1.322	1.465	1.203	1.816	1.269	1.317	1.461	1.461	1.039	1.258	1.125	1.281	1.279	1.240
	Schmidt_FAU_task1_2	Schmidt2023	12	21.50	1.337	1.488	1.210	1.972	1.322	1.321	1.405	1.420	1.038	1.292	1.140	1.274	1.277	1.241
	Schmidt_FAU_task1_3	Schmidt2023	19	26.75	1.398	1.562	1.261	1.961	1.323	1.344	1.620	1.562	1.083	1.332	1.177	1.355	1.329	1.288
	Schmidt_FAU_task1_4	Schmidt2023	24	28.75	1.482	1.650	1.342	1.946	1.407	1.434	1.785	1.677	1.143	1.390	1.314	1.357	1.451	1.395
	Tan_NTU_task1_1	Tan2023	33	33.50	1.508	1.720	1.331	1.789	1.427	1.696	1.879	1.809	1.107	1.387	1.260	1.406	1.446	1.379
	Tan_NTU_task1_2	Tan2023	35	34.00	1.461	1.634	1.317	1.706	1.405	1.630	1.758	1.671	1.097	1.387	1.245	1.400	1.425	1.346
	Tan_NTU_task1_3	Tan2023	37	37.00	1.492	1.632	1.375	1.868	1.461	1.489	1.692	1.650	1.180	1.473	1.302	1.445	1.462	1.388
	Tan_SCUT_task1_1	Tan2023a	4	13.50	1.192	1.420	1.002	2.215	1.096	1.205	1.304	1.279	0.899	1.100	0.952	1.018	1.027	1.019
	Tan_SCUT_task1_2	Tan2023a	30	31.00	1.444	1.726	1.208	2.504	1.262	1.341	1.902	1.619	1.037	1.335	1.164	1.254	1.248	1.214
	Tan_SCUT_task1_3	Tan2023a	16	25.50	1.441	1.675	1.247	2.407	1.278	1.366	1.662	1.661	1.095	1.395	1.192	1.261	1.322	1.214
	Tan_SCUT_task1_4	Tan2023a	32	33.00	1.525	1.843	1.259	2.915	1.307	1.422	1.954	1.620	1.134	1.416	1.195	1.285	1.273	1.253
	Vo_DU_task1_1	Vo2023	52	46.75	2.157	2.963	1.485	6.412	1.811	1.767	2.425	2.399	1.240	1.438	1.346	1.608	1.639	1.641
	Vo_DU_task1_2	Vo2023	53	47.75	2.116	2.877	1.482	6.182	1.769	1.759	2.385	2.292	1.237	1.443	1.349	1.584	1.624	1.652
	Vo_DU_task1_3	Vo2023	50	46.25	2.092	2.827	1.479	5.800	1.808	1.747	2.280	2.501	1.228	1.432	1.318	1.600	1.651	1.647
	Vo_DU_task1_4	Vo2023	49	45.75	1.793	2.169	1.479	2.920	1.752	1.705	2.256	2.214	1.276	1.484	1.376	1.590	1.592	1.556
	Wang_SCUT_task1_1	Wang2023	31	31.50	1.493	1.823	1.218	2.972	1.326	1.431	1.818	1.569	1.056	1.320	1.165	1.228	1.263	1.277
	Wang_SCUT_task1_2	Wang2023	18	26.50	1.348	1.510	1.214	1.798	1.264	1.302	1.632	1.553	1.086	1.309	1.168	1.241	1.235	1.243
	Wang_SCUT_task1_3	Wang2023	46	43.50	1.702	2.228	1.263	4.002	1.475	1.499	1.759	2.405	1.039	1.276	1.159	1.467	1.274	1.366
	Wang_SCUT_task1_4	Wang2023	48	45.00	1.472	1.654	1.321	1.966	1.384	1.415	1.845	1.657	1.126	1.395	1.309	1.341	1.381	1.376
	XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	1.364	1.450	1.292	1.630	1.351	1.345	1.455	1.469	1.220	1.297	1.248	1.317	1.363	1.307
	XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	1.355	1.455	1.271	1.553	1.380	1.347	1.487	1.510	1.222	1.255	1.209	1.300	1.352	1.291
	XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	1.395	1.509	1.301	1.654	1.418	1.391	1.551	1.528	1.225	1.287	1.239	1.340	1.385	1.329
	XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	1.367	1.450	1.298	1.575	1.363	1.371	1.442	1.501	1.215	1.281	1.224	1.331	1.408	1.331
	Yang_GZHU_task1_1	Weng2023	15	24.75	1.280	1.409	1.172	1.705	1.169	1.276	1.423	1.472	0.999	1.254	1.084	1.213	1.234	1.248
	Yang_GZHU_task1_2	Weng2023	14	22.25	1.241	1.352	1.149	1.661	1.141	1.206	1.348	1.403	1.001	1.213	1.077	1.184	1.207	1.215
	Yang_GZHU_task1_3	Weng2023	21	27.25	1.279	1.406	1.172	1.941	1.188	1.204	1.272	1.426	1.017	1.272	1.087	1.175	1.236	1.246
	Yang_GZHU_task1_4	Weng2023	19	26.75	1.259	1.364	1.172	1.794	1.180	1.189	1.260	1.396	1.031	1.241	1.097	1.178	1.243	1.243
	Zhang_NCUT_task1_1	Zhang2023	49	45.75	1.757	2.083	1.485	2.026	1.802	1.898	2.387	2.301	1.178	1.376	1.295	1.655	1.661	1.743
	Zhang_NCUT_task1_2	Zhang2023	51	46.50	1.533	1.813	1.300	1.788	1.506	1.501	2.399	1.869	1.087	1.275	1.127	1.455	1.426	1.431
	Zhang_SATLab_task1_1	Bing2023	26	29.25	3.248	5.085	1.717	9.657	2.122	3.929	7.783	1.935	1.724	1.700	1.609	1.798	1.808	1.662
	Zhang_SATLab_task1_2	Bing2023	34	33.75	4.213	6.653	2.179	12.855	2.905	3.678	10.376	3.450	2.070	2.147	2.139	2.077	2.308	2.335
	Zhang_SATLab_task1_3	Bing2023	27	30.00	1.704	2.237	1.261	4.419	1.475	1.685	2.116	1.487	1.139	1.288	1.124	1.332	1.352	1.329
	Zhang_SATLab_task1_4	Bing2023	22	27.50	1.542	1.802	1.325	2.165	1.581	1.735	2.027	1.504	1.145	1.385	1.206	1.415	1.408	1.392

System characteristics

General characteristics

Submission label	Technical Report	Official system rank	Rank value	Accuracy (Eval)	Logloss (Eval)	Sampling rate	Data augmentation	Features
AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	1.920	8kHz	pitch shifting, time stretching, mixup, time masking, frequency masking	log-mel energies
AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	1.996	8kHz	pitch shifting, time stretching, mixup, time masking, frequency masking	log-mel energies
AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	1.364	8kHz	pitch shifting, time stretching, mixup, time masking, frequency masking	log-mel energies
Bai_JLESS_task1_1	Du2023	54	50.25	47.9	1.825	44.1kHz	FMix	log-mel energies
Bai_JLESS_task1_2	Du2023	47	43.75	48.8	1.791	44.1kHz	FMix	log-mel energies
Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	1.174	32kHz	mixup, mixstyle,	log-mel energies
Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	1.246	32kHz	mixup, mixstyle,	log-mel energies
Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	1.241	32kHz	mixup, mixstyle	log-mel energies
Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	1.252	32kHz	mixup, mixstyle	log-mel energies
Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	1.307	32kHz	device simulation, mixup, mixstyle	log-mel energies
Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	1.292	32kHz	device simulation, mixup, mixstyle	log-mel energies
Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	1.223	32kHz	device simulation, mixup, mixstyle	log-mel energies
Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	1.241	32kHz	device simulation, mixup, mixstyle	log-mel energies
Fei_vv_task1_1	Fei2023	15	24.75	55.2	1.282	16kHz	mixup	log-mel energies
Fei_vv_task1_2	Fei2023	13	21.75	54.5	1.290	16kHz	mixup	log-mel energies
Fei_vv_task1_3	Fei2023	26	29.25	53.2	1.349	16kHz	mixup, time stretching	log-mel energies
Fei_vv_task1_4	Fei2023	23	28.25	51.8	1.370	16kHz	mixup, time stretching	log-mel energies
Han_SZU_task1_1	Han2023	39	37.75	50.5	2.012	44.1kHz	mixup, time masking, frequency masking, pitch shifting, random noise	log-mel energies, spectral envelope, spectrum fine structure
LAM_AEV_task1_1	Pham2023	25	29.00	55.3	1.847	44.1kHz	mixup	CQT, Gammatonegram, Mel
LAM_AEV_task1_2	Pham2023	31	31.50	55.0	2.083	44.1kHz	mixup	CQT, Gammatonegram, Mel
LAM_AEV_task1_3	Pham2023	20	27.00	55.6	1.933	44.1kHz	mixup	CQT, Gammatonegram, Mel
Liang_NTES_task1_1	Liang2023	40	38.50	52.6	1.402	44.1kHz	random cropping, SpecAugment, mixup	log-mel energies
MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	1.230	32kHz	mixstyle	log-mel energies
MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	1.242	32kHz	mixstyle	log-mel energies
MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	4.354	32kHz	mixup	log-mel energies
MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	3.224	32kHz	mixup	log-mel energies
DCASE2023 baseline		52	46.75	44.8	1.523	44.1kHz		log-mel energies
Park_KT_task1_1	Kim2023	10	20.75	56.3	1.495	16kHz	mixup, frequency masking, temporal masking	log-mel energies
Park_KT_task1_2	Kim2023	29	30.75	53.0	1.660	16kHz	mixup, frequency masking, temporal masking	log-mel energies
Park_KT_task1_3	Kim2023	26	29.25	54.2	2.230	16kHz	mixup, frequency masking, temporal masking	log-mel energies
Park_KT_task1_4	Kim2023	20	27.00	49.2	1.510	16kHz	mixup, frequency masking, temporal masking	log-mel energies
Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	1.313	32kHz	device impulse response augmentation, mixup, freq-mixstyle, pitch shifting	log-mel energies
Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	1.256	32kHz	device impulse response augmentation, mixup, freq-mixstyle, pitch shifting	log-mel energies
Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	1.153	32kHz	device impulse response augmentation, mixup, freq-mixstyle, pitch shifting	log-mel energies
Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	1.117	32kHz	device impulse response augmentation, mixup, freq-mixstyle, pitch shifting	log-mel energies
Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	1.322	32kHz	random cutoff, mixstyle, pitch shifting	log-mel energies
Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	1.337	32kHz	random cutoff, mixstyle, pitch shifting	log-mel energies
Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	1.398	32kHz	random cutoff, mixstyle, pitch shifting	log-mel energies
Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	1.482	32kHz	random cutoff, mixstyle, pitch shifting	log-mel energies
Tan_NTU_task1_1	Tan2023	33	33.50	47.1	1.508	44.1kHz		log-mel energies
Tan_NTU_task1_2	Tan2023	35	34.00	48.5	1.461	44.1kHz		log-mel energies
Tan_NTU_task1_3	Tan2023	37	37.00	46.3	1.492	44.1kHz		log-mel energies
Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	1.192	44.1kHz		mel-spectrogram
Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	1.444	44.1kHz		mel-spectrogram
Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	1.441	44.1kHz		mel-spectrogram
Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	1.525	44.1kHz		mel-spectrogram
Vo_DU_task1_1	Vo2023	52	46.75	45.0	2.157	44.1kHz	mixup, SpecAugment	log-mel spectrogram
Vo_DU_task1_2	Vo2023	53	47.75	44.8	2.116	44.1kHz	mixup, SpecAugment	log-mel spectrogram
Vo_DU_task1_3	Vo2023	50	46.25	45.2	2.092	44.1kHz	mixup, SpecAugment	log-mel spectrogram
Vo_DU_task1_4	Vo2023	49	45.75	45.5	1.793	44.1kHz	mixup, SpecAugment	log-mel spectrogram
Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	1.493	44.1kHz	mixup,speculation and spectral modulation	log-mel energies
Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	1.348	44.1kHz	mixup,speculation and spectral modulation	log-mel energies
Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	1.702	44.1kHz	mixup,speculation and spectral modulation	log-mel energies
Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	1.472	44.1kHz	mixup,speculation and spectral modulation	log-mel energies
XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	1.364	32kHz	timerolling, pitch shifting, gaussian noise, specaugment, mixup, mixstyle	log-mel energies
XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	1.355	32kHz	timerolling, pitch shifting, gaussian noise, specaugment, mixup, mixstyle	log-mel energies
XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	1.395	32kHz	timerolling, pitch shifting, gaussian noise, specaugment, mixup, mixstyle	log-mel energies
XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	1.367	32kHz	timerolling, pitch shifting, gaussian noise, specaugment, mixup, mixstyle	log-mel energies
Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	1.280	32kHz	Conv_IR, time shifting, time-frequency masking, mixstyle	log-mel energies
Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	1.241	32kHz	Conv_IR, time shifting, time-frequency masking, mixstyle	log-mel energies
Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	1.279	32kHz	Conv_IR, time shifting, time-frequency masking, mixstyle	log-mel energies
Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	1.259	32kHz	Conv_IR, time shifting, time-frequency masking, mixstyle	log-mel energies
Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	1.757	44.1kHz	mixup, specaugment	log-mel energies,delta and delta-delta
Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	1.533	44.1kHz	mixup, specaugment	log-mel energies,delta and delta-delta
Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	3.248	44.1kHz	mixup, temporal crop, time masking, frequency masking	log-mel energies
Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	4.213	44.1kHz	mixup, temporal crop, time masking, frequency masking	log-mel energies
Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	1.704	44.1kHz	mixup, temporal crop, time masking, frequency masking	log-mel energies
Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	1.542	44.1kHz	mixup, temporal crop, time masking, frequency masking	log-mel energies

Machine learning characteristics

Code	Technical Report	Official system rank	Rank value	Accuracy (Eval)	Logloss (Eval)	External data usage	External data sources	Model complexity	Model MACS	Classifier	Ensemble subsystems	Decision making	Framework	Pipeline
AI4EDGE_IPL_task1_1	Almeida2023	28	30.25	51.9	1.920			52852	25475456	CNN, ensemble	10		keras/tensorflow	pretraining, training, adaptation, knowledge distillation, weight quantization
AI4EDGE_IPL_task1_2	Almeida2023	47	43.75	48.8	1.996			68996	29304736	CNN, ensemble	10		keras/tensorflow	pretraining, training, adaptation, knowledge distillation, weight quantization
AI4EDGE_IPL_task1_3	Almeida2023	38	37.25	50.8	1.364			65192	26711936	CNN, ensemble	10		keras/tensorflow	pretraining, training, adaptation, knowledge distillation, weight quantization
Bai_JLESS_task1_1	Du2023	54	50.25	47.9	1.825			78252	27931612	CNN,ResNet,Transformer,CBAM			pytorch	pretraining, training, adaptation, pruning, weight quantization
Bai_JLESS_task1_2	Du2023	47	43.75	48.8	1.791			60458	14130372	CNN,ResNet,Transformer,CBAM			pytorch	pretraining, training, adaptation, pruning, weight quantization
Cai_TENCENT_task1_1	Cai2023	17	25.75	56.6	1.174	pre-trained model	Audioset	127684	28840396	CNN			pytorch	pretraining, teacher model trainining, student model training, weight quantization
Cai_TENCENT_task1_2	Cai2023	11	21.25	56.2	1.246	pre-trained model	Audioset	79942	21990724	CNN			pytorch	pretraining, teacher model trainining, student model training, weight quantization
Cai_TENCENT_task1_3	Cai2023	14	22.25	55.8	1.241	pre-trained model	Audioset	79942	21990724	CNN			pytorch	pretraining, teacher model trainining, student model training, weight quantization
Cai_TENCENT_task1_4	Cai2023	11	21.25	55.4	1.252	pre-trained model	Audioset	63558	19533124	CNN			pytorch	pretraining, teacher model trainining, student model training, weight quantization
Cai_XJTLU_task1_1	Cai2023a	9	19.75	51.9	1.307	device simulation	MicIRP	6828	1649349	CNN, TF-SepNet			pytorch	training, weight quantization
Cai_XJTLU_task1_2	Cai2023a	8	18.25	52.5	1.292	device simulation	MicIRP	6828	1649349	CNN, TF-SepNet			pytorch	training, weight quantization
Cai_XJTLU_task1_3	Cai2023a	6	14.00	55.1	1.223	device simulation	MicIRP	15890	3424245	CNN, TF-SepNet			pytorch	training, weight quantization
Cai_XJTLU_task1_4	Cai2023a	3	11.50	57.0	1.241	device simulation	MicIRP	54260	10219540	CNN, TF-SepNet			pytorch	training, weight quantization
Fei_vv_task1_1	Fei2023	15	24.75	55.2	1.282	pre-trained model		123636	13402932	SERFR-CNN-32			pytorch	pretraining, training, adaptation, pruning, weight quantization
Fei_vv_task1_2	Fei2023	13	21.75	54.5	1.290	pre-trained model		70588	7802348	SERFR-CNN-24			pytorch	pretraining, training, adaptation, pruning, weight quantization
Fei_vv_task1_3	Fei2023	26	29.25	53.2	1.349	pre-trained model		123636	13402932	SERFR-CNN-32			pytorch	pretraining, training, adaptation, pruning, weight quantization
Fei_vv_task1_4	Fei2023	23	28.25	51.8	1.370	pre-trained model		70588	7802348	SERFR-CNN-24			pytorch	pretraining, training, adaptation, pruning, weight quantization
Han_SZU_task1_1	Han2023	39	37.75	50.5	2.012			80845	29.349M	CNN			keras/tensorflow	cepstrum analysis, extract features, train teacher_network, knowledge distillation, train student_network
LAM_AEV_task1_1	Pham2023	25	29.00	55.3	1.847			22962	29267550	CNN	3	Late fusion of predicted probabilities	keras/tensorflow	training
LAM_AEV_task1_2	Pham2023	31	31.50	55.0	2.083			22962	29267550	CNN	3	Late fusion of predicted probabilities	keras/tensorflow	training
LAM_AEV_task1_3	Pham2023	20	27.00	55.6	1.933			22962	29267550	CNN	3	Late fusion of predicted probabilities	keras/tensorflow	training
Liang_NTES_task1_1	Liang2023	40	38.50	52.6	1.402			31260	29591778	CNN			pytorch	training teacher, training student
MALACH23_JKU_task1_1	Pichler2023	8	18.25	57.0	1.230			59804	14686940	CNN		maximum likelihood	pytorch	mel spectrogram, mixstyle, training
MALACH23_JKU_task1_2	Pichler2023	7	16.75	56.6	1.242			43580	10819292	CNN		maximum likelihood	pytorch	mel spectrogram, mixstyle, training
MALACH23_JKU_task1_3	Pichler2023	42	39.50	9.9	4.354			116648	572340	S4		maximum likelihood	pytorch	mel spectrogram, mixup, training
MALACH23_JKU_task1_4	Pichler2023	43	39.75	9.8	3.224			15994	214420	S4		maximum likelihood	pytorch	mel spectrogram, mixup, training
DCASE2023 baseline		52	46.75	44.8	1.523	embeddings		46512	29234920	CNN			keras/tensorflow	pretraining, training, adaptation, pruning, weight quantization
Park_KT_task1_1	Kim2023	10	20.75	56.3	1.495	embeddings		92070	19556096	BCRes2Net		maximum likelihood	pytorch	training, weight quantization
Park_KT_task1_2	Kim2023	29	30.75	53.0	1.660	embeddings		92070	19556096	BCRes2Net		maximum likelihood	pytorch	training, weight quantization
Park_KT_task1_3	Kim2023	26	29.25	54.2	2.230	embeddings		92070	19556096	BCRes2Net		maximum likelihood	pytorch	training, weight quantization
Park_KT_task1_4	Kim2023	20	27.00	49.2	1.510	embeddings		20516	617000	BCRes2Net		maximum likelihood	pytorch	training, weight quantization
Schmid_CPJKU_task1_1	Schmid2023	5	13.75	54.4	1.313	pre-trained model	PaSST, MicIRP	5722	1582336	RF-regularized CNNs, PaSST transformer			pytorch	train teachers, ensemble teacher logits, train student using knowledge distillation, quantization-aware training
Schmid_CPJKU_task1_2	Schmid2023	1	5.25	58.7	1.256	pre-trained model	PaSST, MicIRP	12310	4354304	RF-regularized CNNs, PaSST transformer			pytorch	train teachers, ensemble teacher logits, train student using knowledge distillation, quantization-aware training
Schmid_CPJKU_task1_3	Schmid2023	2	7.00	61.4	1.153	pre-trained model	PaSST, MicIRP	30106	9638144	RF-regularized CNNs, PaSST transformer			pytorch	train teachers, ensemble teacher logits, train student using knowledge distillation, quantization-aware training
Schmid_CPJKU_task1_4	Schmid2023	3	11.50	62.7	1.117	pre-trained model	PaSST, MicIRP	54182	16803072	RF-regularized CNNs, PaSST transformer			pytorch	train teachers, ensemble teacher logits, train student using knowledge distillation, pruning, quantization-aware training
Schmidt_FAU_task1_1	Schmidt2023	26	29.25	55.7	1.322	pre-trained model		127988	28931380	ensemble, RF-regularized CNNs, PaSST transformer	8	generalized mean	pytorch	pretraining, training, adaptation, pruning, weight quantization
Schmidt_FAU_task1_2	Schmidt2023	12	21.50	55.6	1.337	pre-trained model		68456	19910080	ensemble, RF-regularized CNNs, PaSST transformer	8	weighted generalized mean	pytorch	pretraining, training, adaptation, pruning, weight quantization
Schmidt_FAU_task1_3	Schmidt2023	19	26.75	52.7	1.398	pre-trained model		74700	9996775	ensemble, RF-regularized CNNs, PaSST transformer	8	generalized mean	pytorch	pretraining, training, adaptation, pruning, weight quantization
Schmidt_FAU_task1_4	Schmidt2023	24	28.75	49.7	1.482	pre-trained model		34616	4938255	ensemble, RF-regularized CNNs, PaSST transformer	8	generalized mean	pytorch	pretraining, training, adaptation, pruning, weight quantization
Tan_NTU_task1_1	Tan2023	33	33.50	47.1	1.508			37434	2960384	CNN			keras/tensorflow	training, weight quantization
Tan_NTU_task1_2	Tan2023	35	34.00	48.5	1.461			54242	6462656	CNN			keras/tensorflow	training, weight quantization
Tan_NTU_task1_3	Tan2023	37	37.00	46.3	1.492			54242	6462656	CNN			keras/tensorflow	training, weight quantization
Tan_SCUT_task1_1	Tan2023a	4	13.50	60.8	1.192	embeddings		73386	13180000	CNN			pytorch	training teacher, training student, knowledge distillation, weight quantization
Tan_SCUT_task1_2	Tan2023a	30	31.00	51.7	1.444	embeddings		73386	13180000	CNN			pytorch	training teacher, training student, knowledge distillation, weight quantization
Tan_SCUT_task1_3	Tan2023a	16	25.50	53.5	1.441	embeddings		73386	13180000	CNN			pytorch	training teacher, training student, knowledge distillation, weight quantization
Tan_SCUT_task1_4	Tan2023a	32	33.00	50.9	1.525	embeddings		73386	13180000	CNN			pytorch	training teacher, training student, knowledge distillation, weight quantization
Vo_DU_task1_1	Vo2023	52	46.75	45.0	2.157	pre-trained model		119526	15600000	CNN, Transformer, Knowledge Distillation			pytorch	pretraining, training, adaptation, weight quantization
Vo_DU_task1_2	Vo2023	53	47.75	44.8	2.116	pre-trained model		119526	15600000	CNN, Transformer, Knowledge Distillation			pytorch	pretraining, training, adaptation, weight quantization
Vo_DU_task1_3	Vo2023	50	46.25	45.2	2.092	pre-trained model		119526	15600000	CNN, Transformer, Knowledge Distillation			pytorch	pretraining, training, adaptation, weight quantization
Vo_DU_task1_4	Vo2023	49	45.75	45.5	1.793	pre-trained model		119526	15600000	CNN, Transformer, Knowledge Distillation			pytorch	pretraining, training, adaptation, weight quantization
Wang_SCUT_task1_1	Wang2023	31	31.50	49.1	1.493			45164	8646000	CNN			pytorch	training, adaptation, pruning, weight quantization
Wang_SCUT_task1_2	Wang2023	18	26.50	52.9	1.348			56172	16746000	CNN			pytorch	training, adaptation, pruning, weight quantization
Wang_SCUT_task1_3	Wang2023	46	43.50	47.1	1.702			56556	25442000	CNN			pytorch	training, adaptation, pruning, weight quantization
Wang_SCUT_task1_4	Wang2023	48	45.00	48.5	1.472			121812	20902000	CNN			pytorch	training, adaptation, pruning, weight quantization
XuQianHu_BIT&NUDT_task1_1	Yu2023	45	40.50	51.0	1.364	pre-trained model		52288	23803968	CNN + Transformer			pytorch	pretraining, training, adaptation, distillation, weight quantization
XuQianHu_BIT&NUDT_task1_2	Yu2023	44	40.00	51.6	1.355	pre-trained model		51648	28400320	CNN + Transformer			pytorch	pretraining, training, adaptation, distillation, weight quantization
XuQianHu_BIT&NUDT_task1_3	Yu2023	41	39.25	50.0	1.395	pre-trained model		57392	13402688	CNN + Transformer			pytorch	pretraining, training, adaptation, distillation, weight quantization
XuQianHu_BIT&NUDT_task1_4	Yu2023	36	35.50	51.1	1.367	pre-trained model		66114	11878580	CNN + Transformer			pytorch	pretraining, training, adaptation, distillation, weight quantization
Yang_GZHU_task1_1	Weng2023	15	24.75	55.5	1.280	pre-trained model, convolution with IRs from MicIRP	Microphone Impulse Response Project	76906	23970000	CNN			pytorch	pretraining, DML training, KD fine-tuning, weight quantization
Yang_GZHU_task1_2	Weng2023	14	22.25	55.9	1.241	pre-trained model, convolution with IRs from MicIRP	Microphone Impulse Response Project	76906	23970000	CNN		average	pytorch	pretraining, DML training, KD fine-tuning, weight quantization
Yang_GZHU_task1_3	Weng2023	21	27.25	55.1	1.279	pre-trained model, convolution with IRs from MicIRP	Microphone Impulse Response Project	76906	23970000	CNN			pytorch	pretraining, DML training, weight quantization
Yang_GZHU_task1_4	Weng2023	19	26.75	55.2	1.259	pre-trained model, convolution with IRs from MicIRP	Microphone Impulse Response Project	76906	23970000	CNN		average	pytorch	pretraining, DML training, weight quantization
Zhang_NCUT_task1_1	Zhang2023	49	45.75	43.3	1.757		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	123648	7375000	GhostNet		average	keras/tensorflow	training, weight quantization
Zhang_NCUT_task1_2	Zhang2023	51	46.50	47.9	1.533		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	76224	28461000	FHR_Mobilenet		average	keras/tensorflow	training, weight quantization
Zhang_SATLab_task1_1	Bing2023	26	29.25	48.8	3.248		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	7946	3972096	MobileNet			keras/tensorflow	training, weight quantization
Zhang_SATLab_task1_2	Bing2023	34	33.75	50.0	4.213		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	46232	19466240	mini-SegNet			keras/tensorflow	training, pruning, retraining, weight quantization
Zhang_SATLab_task1_3	Bing2023	27	30.00	51.9	1.704		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	54178	23438336	ensemble, MobileNet, mini-SegNet	2		keras/tensorflow	training, pruning, weight quantization, ensemble
Zhang_SATLab_task1_4	Bing2023	22	27.50	50.3	1.542		TAU Urban Acoustic Scenes 2022 Mobile, Development dataset	15892	7944192	MobileNet, ensemble	2		keras/tensorflow	training, weight quantization, ensemble

Technical reports

Ai4edgept Submission to DCASE 2023 Low Complexity Acoustic Scene Classification Task1

Carlos Almeida¹, Piovesan Federico², Luis Bento¹ and Mónica Figueiredo¹

¹Eletrotechnical Engineering, Instituto Politécnico de Leiria, Leiria, Portugal, ²Eletrotechnical Engineering, Politecnico di Torino, Torino, Italy

AI4EDGE_IPL_task1_1 AI4EDGE_IPL_task1_2 AI4EDGE_IPL_task1_3

PDF Code

Ai4edgept Submission to DCASE 2023 Low Complexity Acoustic Scene Classification Task1

Carlos Almeida¹, Piovesan Federico², Luis Bento¹ and Mónica Figueiredo¹
¹Eletrotechnical Engineering, Instituto Politécnico de Leiria, Leiria, Portugal, ²Eletrotechnical Engineering, Politecnico di Torino, Torino, Italy

Abstract

The DCASE task 1 challenge aims to classify acoustic scenes using devices with low computational power and memory. The DCASE2023 challenge gives further importance to the size and multiply-accumulate operation count (MAC), this report aims to describe the submission to this challenge, following our research group’s previous work in this field, and the model submitted to DCASE 2022. We use a one-versus-all ten-network ensemble model and propose a knowledge distillation custom method to reduce model complexity. The ensemble model is used as the teacher network, distilling knowledge to the student. The student has 3 variations, the first model is a tuned version of the DCASE2022 baseline architecture, for the second model a slightly larger version of the first model and for the third model a larger version of the second model using structured pruning to further reduce model complexity. Data preprocessing is also conducted in order to further improve performance. Results show that the proposed knowledge distillation methods were able to improve the accuracy significantly.

System characteristics

Sampling rate	8kHz
Data augmentation	pitch shifting, time stretching, mixup, time masking, frequency masking
Features	log-mel energies
Classifier	CNN, ensemble
Complexity management	weight quantization; weight quantization, pruning

Content

Task description

Systems ranking

Teams ranking

System complexity

Energy Consumption

Generalization performance

Class-wise performance

Accuracy

Log loss

Device-wise performance

Accuracy

Log loss

System characteristics

General characteristics

Machine learning characteristics

Technical reports

Ai4edgept Submission to DCASE 2023 Low Complexity Acoustic Scene Classification Task1

Ai4edgept Submission to DCASE 2023 Low Complexity Acoustic Scene Classification Task1

Abstract

System characteristics

Mini-Segnet and Low-Complexity Mobilenet for Acoustic Scene Classification

Mini-Segnet and Low-Complexity Mobilenet for Acoustic Scene Classification

Abstract

System characteristics

Tencent Submission to Dcase23 Task1: Low-Complexity Deep Learning Solution for Acoustic Scene Classification

Tencent Submission to Dcase23 Task1: Low-Complexity Deep Learning Solution for Acoustic Scene Classification

Abstract

System characteristics

Dcase2023 Task1 Submission: Device Simulation and Time-Frequency Separable Convolution for Acoustic Scene Classification

Dcase2023 Task1 Submission: Device Simulation and Time-Frequency Separable Convolution for Acoustic Scene Classification

Abstract

System characteristics

How Information on Soft Labels and Hard Labels Mutually Benefits Sound Event Detection Tasks

How Information on Soft Labels and Hard Labels Mutually Benefits Sound Event Detection Tasks

Abstract

System characteristics

Acoustic Scene Classification Based on Multi-Teacher Knowledge Distillation and Serfr-CNN

Acoustic Scene Classification Based on Multi-Teacher Knowledge Distillation and Serfr-CNN

Abstract

System characteristics

Submission to DCASE 2023 Task 1: Low-Complexity Acoustic Scene Classification Using Cepstral Analysis

Submission to DCASE 2023 Task 1: Low-Complexity Acoustic Scene Classification Using Cepstral Analysis

Abstract

System characteristics

Dual-Strategy Enhancement of Acoustic Scene and Event Classification: Integrating Res2net, Ghostnet, and Mobileformer Architectures

Dual-Strategy Enhancement of Acoustic Scene and Event Classification: Integrating Res2net, Ghostnet, and Mobileformer Architectures

Abstract

System characteristics

Low-Complexity Acoustic Scene Classification Base on Depthwise Separable CNN

Low-Complexity Acoustic Scene Classification Base on Depthwise Separable CNN

Abstract

System characteristics

Low-Complexity Deep Learning System for Acoustic Scene Classification Using Teacher-Student Scheme and Multiple Spectrograms

Low-Complexity Deep Learning System for Acoustic Scene Classification Using Teacher-Student Scheme and Multiple Spectrograms

Abstract

System characteristics

Malach23 Submission to DCASE 2023: Acoustic Scene Classification with Receptive-Field Regularized Convolution Neural Networks and State Space Models

Malach23 Submission to DCASE 2023: Acoustic Scene Classification with Receptive-Field Regularized Convolution Neural Networks and State Space Models

Abstract

System characteristics

CP-JKU Submission to Dcase23: Efficient Acoustic Scene Classification with Cp-Mobile

CP-JKU Submission to Dcase23: Efficient Acoustic Scene Classification with Cp-Mobile

Abstract

System characteristics

Submission to DCASE 2023 Task 1: Device Invariant Training with Structured Filter Pruning for Low Complexity Acoustic Scene Classification

Submission to DCASE 2023 Task 1: Device Invariant Training with Structured Filter Pruning for Low Complexity Acoustic Scene Classification

Abstract

System characteristics

Low-Complexity Acoustic Scene Classification Using Convolution Neural Network

Low-Complexity Acoustic Scene Classification Using Convolution Neural Network

Abstract

System characteristics

Low-Complexity Acoustic Scene Classification Using Blueprint Separable Convolution and Knowledge Distillation

Low-Complexity Acoustic Scene Classification Using Blueprint Separable Convolution and Knowledge Distillation

Abstract

System characteristics

Hierarchical Knowledge Distillation: A Multi-Stage Learning Approach

Hierarchical Knowledge Distillation: A Multi-Stage Learning Approach

Abstract