5-way 5-shot incremental learning on miniImageNet (TOPIC) benchmark leaderboard · SOTA2 Research

5-way 5-shot incremental learning on miniImageNet (TOPIC)

72.3Accuracy (Session 1)Rebalance

Evaluation Results

Method	Links
Rebalance2021.10		72.3	66.37	61	56.93	53.31	49.93	46.47	44.13	42.19	-6.12
CRT2021.10		72.08	68.15	63.06	61.12	56.57	54.47	51.81	49.86	48.31	—
Joint-training2021.10		72.08	67.31	62.04	58.51	54.41	51.53	48.7	45.49	43.88	-4.43
Baseline2021.10		72.08	66.29	61.99	58.71	55.73	53.04	50.4	48.59	47.31	-1
FSLL2021.10		72.08	59.04	53.75	51.17	49.11	47.21	45.35	44.06	43.65	-4.66
F2M2021.10		72.05	67.47	63.16	59.7	56.71	53.77	51.11	49.21	47.84	-0.43
iCaRL2021.10		71.77	61.85	58.12	54.6	51.49	48.47	45.9	44.19	42.71	-5.6
FSLL+SS2021.10		68.85	63.14	59.24	55.23	52.24	49.65	47.74	45.23	43.92	-4.39
FSLL2021.10		66.48	61.75	58.16	54.16	51.1	48.53	46.54	44.2	42.28	-6.03
iCaRL2021.10		61.31	46.32	42.94	37.63	30.49	24	20.89	18.8	17.21	-31.1
Rebalance2021.10		61.31	47.8	39.31	31.91	25.68	21.35	18.67	17.24	14.17	-34.14
TOPIC2021.10		61.31	50.09	45.17	41.16	37.48	35.52	32.19	29.46	24.42	-23.89