Fine-Grained Visual Recognition on Open-world FGVR (Seen Categories)

73.53Accuracy (Air)Fine-R1-7B

Evaluation Results

Method
Fine-R1-7B2026.02	73.53	86.12	90.73	80.71	81.46	83.14	82.62
Fine-R1-7B2026.02	63.44	75.22	78.88	78.41	86.92	86.76	78.27
Qwen2.5-VL-7B2026.02	58.86	65.97	56.94	59.02	62.61	35.59	56.5
Qwen2.5-VL-3B2026.02	56.98	66.77	52.49	65.12	68.96	26.27	56.1
Qwen2-VL-7B2026.02	56.47	56.46	55.31	67.03	75.02	36.97	57.88
Fine-R1-3B2026.02	54.36	78.9	82.46	78.21	64.55	81.6	73.35
Qwen2-VL-2B2026.02	47.49	48.72	52.95	51.22	66.56	19.88	47.8
Fine-R1-3B2026.02	47.3	66.09	71.15	73.45	77.16	82.58	69.62
Qwen2.5-VL-7B2026.02	46.85	58.28	67.14	68.9	73.09	33.55	57.97
Qwen2-VL-7B2026.02	45.5	37.93	66.16	53.32	69.81	27.48	50.03
DeepPerception-7B2026.02	44.24	47.63	54.14	49.16	47.3	40.9	47.23
DeepPerception-7B2026.02	40.66	47.63	66.62	64.78	75.47	67.37	60.42
Qwen2.5-VL-3B2026.02	37.96	48.78	58.08	51.19	61.15	13.92	45.18
Qwen2-VL-2B2026.02	34.37	25.04	60.42	41.32	59.06	4.5	37.45
InternVL2.5-8B2026.02	23.76	28.44	30.08	27.11	21.73	29.03	26.69
InternVL2.5-4B2026.02	14.71	25.41	32.19	33.13	23.84	28.19	26.25
InternVL2.5-8B2026.02	10.04	10.97	12.78	18.76	26.34	23.12	17
LLaVA-Onevision-7B2026.02	9.9	31.74	21.35	30.13	45.24	17.87	26.04
Idefics3-LLaMA3-8B2026.02	9.66	27.72	22.96	35.39	40.92	20.17	26.14
InternVL2.5-4B2026.02	8.59	7.67	16.98	19.92	25.12	16.91	15.87
Idefics2-8B2026.02	7.49	11.02	17.94	18.01	54.27	11.12	19.98
InternVL2.5-2B2026.02	7.26	21.32	27.29	25.87	23.08	26.56	21.9
LLaVA-Onevision-7B2026.02	6.44	9.27	21.33	22.55	46.5	3.26	18.23
InternVL2.5-2B2026.02	3.9	6.03	10.37	17.13	25.79	20.5	13.95
Idefics2-8B2026.02	3.64	19.68	19.54	10.03	14.94	2.5	11.72
Idefics3-LLaMA3-8B2026.02	3.3	4.79	5.57	17.04	32.37	3.72	11.13
LLaVA-v1.6-mistral-7B2026.02	2.73	16.02	21.75	19.35	10.33	12.47	13.78
LLaVA-v1.6-mistral-7B2026.02	2	3.27	9.85	16.05	24.75	13.97	11.65