Generalized Zero-Shot Learning on AWA1 benchmark leaderboard · SOTA2 Research

Generalized Zero-Shot Learning on AWA1

91.3S ScoreRelationNet

Evaluation Results

Method
RelationNet2019.04	91.3	31.4	46.7
CONSE2019.10	88.6	0.4	0.8
CMT2018.12	87.6	0.9	1.8
SYNC2018.12	87.3	8.9	16.2
SYNC2019.04	87.3	8.9	16.2
DEM2019.04	84.7	32.8	47.3
TAFE-Net2019.04	84.4	50.5	63.2
CV +mem2022.03	81.35	39.63	53.29
IZF2021.03	80.5	61.3	69.6
Unified feature-generative framework2022.03	78.99	69.33	73.33
CA +mem2022.03	78.56	37.41	50.43
SAE2019.10	77.1	1.8	3.5
Li et al.2021.03	77	62.7	69.1
TCN2021.03	76.5	49.4	60
LisGAN2021.03	76.3	52.6	62.3
ALE2018.12	76.1	16.8	27.5
ALE2019.10	76.1	16.8	27.5
ALE2019.04	76.1	16.8	27.5
EZSL2018.12	75.6	6.6	12.1
ESZSL2019.10	75.6	6.6	12.1
ESZSL2019.04	75.6	6.6	12.1
Seq-CADA2022.03	75.45	23.28	45.58
RFF-GZSL2021.03	75.1	59.8	66.5
TF-VAEGAN2021.03	75.1	59.8	66.6
SJE2018.12	74.6	11.3	19.6
SJE2019.04	74.6	11.3	19.6
DVGR-CZSL2022.03	73.55	44.21	55.22
CE-GZSL2021.03	73.4	65.3	69.1
CADA-VAE2018.12	72.8	57.3	64.1
CADA-VAE2021.03	72.8	57.3	64.1
Seq-CVAE2022.03	72.41	26.39	38.68
LATEM2018.12	71.7	7.3	13.3
LATEM2019.04	71.7	7.3	13.3
DeViSE2018.12	68.7	13.4	22.4
DEVISE2019.10	68.7	13.4	22.4
DeViSE2019.04	68.7	13.4	22.4
SE2018.12	67.8	56.3	61.5
SE-GZSL2021.03	67.8	56.3	61.5
Zhu et al.2021.03	67.1	57.3	61.8
Tf-GCZSL2022.03	65.01	50.43	56.8
cycle-CLSWGAN2021.03	64	56.9	60.2
f-CLSWGAN2018.12	61.4	57.9	59.6
f-CLSWGAN2021.03	61.4	57.9	59.6
f-CLSWGAN2019.04	57.9	61.4	59.6
SE2019.04	56.3	67.8	61.5
EWC2022.03	53.62	—	—
CJME2019.10	47.9	29.8	36.7
ReViSE2018.12	37.1	46.1	41.1
Seq-fCLSWGAN2022.03	24.83	15.29	18.93
CVAE2018.12	—	—	47.2