Camera Movement Understanding on ACaM synthetic videos (test) benchmark leaderboard · SOTA2 Research

Camera Movement Understanding on ACaM synthetic videos (test)

99.44Static AccuracyVLM-3R

Evaluation Results

Method	Links
VLM-3R2026.07		99.44	18.9	43.45	51.92	47.27	89.29	48.68	38.94
Human Performance2026.07		98.88	97.72	98.04	96.15	100	97.62	98.05	97.61
Qwen3-VL2026.07		97.77	54.42	42.92	73.08	40.74	72.62	58.02	53.68
LLaVa-OV2026.07		96.09	39.6	37.25	36.54	51.85	86.9	51.06	41.59
InternVL3.52026.07		96.09	47.29	33.99	67.31	35.19	73.81	51.74	47.73
ShotVL2026.07		96.09	69.8	63.62	63.46	29.63	98.81	71.33	64.49
Mega-SaM2026.07		95.53	76.35	61.66	—	—	—	—	—
InternVideo2.52026.07		94.97	44.73	31.15	75	27.78	91.67	50.98	46.71
Our SFT Qwen3-VL-8B2026.07		94.41	72.36	74.51	82.69	68.52	91.67	78.2	77.44
Gemini-3.1-Pro2026.07		92.07	62	74.07	69.23	55.56	85.37	73.01	69.58
Qwen3-VL2026.07		91.62	68.95	55.77	82.69	51.85	67.86	67.01	64.17
Qwen2.5-VL2026.07		91.06	52.99	50.54	82.69	61.11	94.05	62.43	58.5
Qwen3-VL2026.07		90.5	62.39	45.1	80.77	55.56	83.33	61.92	60.91
Gemini-3-Pro2026.07		89.94	61.54	68.63	73.08	53.7	88.1	70.65	66.27
InternVL3.52026.07		88.27	41.6	47.93	76.92	31.48	86.9	55.47	51.15
CamReasoner2026.07		84.92	51.28	42.7	67.31	55.56	77.38	55.81	53.74
Our SFT Qwen3-VL-4B2026.07		82.68	74.36	66.67	78.85	68.52	84.52	73.28	74.4
Spatial-MLLM2026.07		68.33	29.65	27.07	50	43.64	94.05	40.75	36.04
ViPE2026.07		68.16	52.42	78.65	—	—	—	—	—
GPT-52026.07		64.8	69.8	57.95	82.69	53.7	63.1	63.78	65.09
CameraModel2026.07		58.1	62.68	51.63	71.15	88.89	98.81	61.83	68.45
Random Guess2026.07		25	25	25	25	25	25	25	25
G2VLM2026.07		4.44	2.91	6.99	1.92	5.45	13.1	5.54	4.51