Share your thoughts, 1 month free Claude Pro on usSee more

Video Temporal Grounding on ActivityNet-Captions

53.67Recall @ IoU=0.5

UniTime-7B

Updated 12d ago

Evaluation Results

Method
UniTime-7B 2026.04	53.67	-	35.9	-	44.79	-	-	-	-
UniversalVTG 2026.04	48.98	-	29.79	-	39.39	-	-	-	-
SnAG 2026.04	48.55	-	30.56	-	39.56	-	-	-	-
Factum-4B 2026.04	48.4	69.8	28.1	-	-	-	-	-	-
UnLoc-L 2026.04	48.3	-	30.2	-	39.25	-	-	-	-
MMN 2025.03	48.2	64.5	29.4	46.6	-	-	-	-	-
VDI 2025.03	48.1	-	28.8	-	-	-	-	-	-
2D-TAN 2026.04	46.16	-	29.21	-	37.69	-	-	-	-
Qwen3-VL-8B + F2G-FT 2026.05	46.1	63.8	28.4	45.7	-	-	-	-	-
ED-VTG-7B 2026.04	45.1	-	22.7	-	33.9	-	-	-	-
2D-TAN 2025.03	43.4	60.4	25	42.5	-	-	-	-	-
UniVTG 2026.04	42.41	-	21.55	-	31.98	-	-	-	-
Qwen3-VL-8B + FT 2026.05	39.2	58.4	21.7	40.8	-	-	-	-	-
Time-R1-7B 2026.04	39	58.1	21.4	-	-	-	-	-	-
TRACE-7B 2026.04	37.7	54	24	-	-	-	-	-	-
NumPro 2026.05	37.5	55.6	20.6	38.8	-	-	-	-	-
Qwen3-VL-4B-Instruct 2026.04	35.8	50.2	21.6	-	-	-	-	-	-
videollama (w. TCAS) 2025.10	35.2	-	20	-	-	-	-	-	-
Gemini-2.5-Flash 2026.04	34.7	51.2	21	-	-	-	-	-	-
HawkEye-7B 2026.04	34.7	-	17.9	-	26.3	-	-	-	-
videollama 2025.10	34.3	-	19.1	-	-	-	-	-	-
VideoChat-R1-7B 2026.04	33.3	-	16.7	-	-	-	-	-	-
Gemini-2.0-Flash 2026.04	33.2	50.4	19.9	-	-	-	-	-	-
Qwen3-VL-4B-Thinking 2026.04	31.7	47.8	19	-	-	-	-	-	-
VideoMind 2025.03	30.3	48.4	15.7	33.3	-	-	-	-	-
VTimeLLM 2026.05	29.5	44.8	14.2	31.4	-	-	-	-	-
HawkEye 2026.05	29.3	49.1	10.7	32.7	-	-	-	-	-
Qwen3-VL-8B 2026.05	27.8	42.8	17.3	32.2	-	-	-	-	-
Momentor 2026.05	26.6	42.6	11.6	28.5	-	-	-	-	-
VideoMind 2025.03	26.5	44	12.6	30.1	-	-	-	-	-
LITA 2026.05	25.9	-	-	28.6	-	-	-	-	-
Momentor 2025.03	23	42.9	12.4	29.3	-	-	-	-	-
ChatVTG 2025.03	22.5	40.7	9.4	27.2	-	-	-	-	-
Qwen2.5-VL-7B 2026.04	20.8	34.5	11.2	-	-	-	-	-	-
Qwen2.5-VL-7B 2026.04	16.96	-	8.75	-	12.86	-	-	-	-
TimeChat 2026.05	16.9	30.2	8.2	21.8	-	-	-	-	-
Qwen2.5vl 2025.10	16.5	-	7.8	-	-	-	-	-	-
Qwen2.5vl (w. TCAS) 2025.10	16.5	-	7.9	-	-	-	-	-	-
Valley 2025.03	13.7	30.6	8.1	21.9	-	-	-	-	-
Video-ChatGPT 2025.03	13.6	26.4	6.1	18.9	-	-	-	-	-
E.T. Chat 2025.03	12.8	24.1	6.1	18.9	-	-	-	-	-
VideoChat 2025.03	3.7	8.8	1.5	7.2	-	-	-	-	-
Video-LLaMA 2025.03	2.1	6.9	0.8	6.5	-	-	-	-	-
VLG-NET 2024.12	-	-	-	-	-	46.32	29.82	77.15	63.33
MGSL-Net 2024.12	-	-	-	-	-	51.87	31.42	82.6	66.71
MMN 2024.12	-	-	-	-	-	48.59	29.26	79.5	64.76
SSRN 2024.12	-	-	-	-	-	54.49	33.15	84.72	68.48
G2L 2024.12	-	-	-	-	-	51.68	33.35	81.32	67.6
Contrastive-MSAT 2024.12	-	-	-	-	-	47.73	31.21	78.06	63.63
SnAG 2024.12	-	-	-	-	-	48.55	30.56	81.71	63.41
Multi-Scale Contrastive Learning 2024.12	-	-	-	-	-	54.83	33.56	84.78	68.91
TimeChat 2026.05	-	36.2	-	21.8	-	20.2	9.5	-	-
HawkEye 2026.05	-	49.1	-	-	-	29.3	10.7	-	-
VTimeLLM 2026.05	-	44	-	30.4	-	27.8	14.3	-	-
Momentor 2026.05	-	42.9	-	29.3	-	23	12.4	-	-
Qwen2.5-VL* 2026.05	-	38.7	-	28.6	-	25.6	14.9	-	-
Grounded-VideoLLM 2026.05	-	43.9	-	34.5	-	29.1	18.3	-	-
ED-VTG 2026.05	-	52.1	-	35.2	-	33.1	16	-	-
VideoChat-R1 2026.05	-	50.4	-	34.3	-	32.2	16.2	-	-
VideoChat-R1.5 2026.05	-	52.4	-	35.5	-	32.3	16.8	-	-
Time-R1 2026.05	-	58.6	-	40.5	-	39	21.4	-	-
EvoGround 2026.05	-	62.9	-	43.9	-	43.6	25	-	-