Spatio-temporal Video Grounding on VidSTG Declarative Sentences (test) benchmark leaderboard · SOTA2 Research

Spatio-temporal Video Grounding on VidSTG Declarative Sentences (test)

37.2m_vIoUBridge-STG

Evaluation Results

Method	Links
Bridge-STG2026.04		37.2	52.6	52.4	37.4
TA-STVG2026.04		34.4	51.7	48.2	33.5
CG-STVG2026.04		34	51.4	47.7	33.1
STCAT2022.09		33.14	50.82	46.2	32.58
TubeDETR2022.03		30.4	48.1	42.5	28.2
TubeDETR2022.09		30.4	48.1	42.5	28.2
TubeDETR2026.04		30.4	48.1	42.5	28.2
SpaceVLLM2026.04		27.4	47.7	39.1	26.2
STVGBert2022.03		24	—	30.9	18.4
STVGBert2022.09		23.97	—	30.91	18.39
OMRN2022.09		23.11	50.73	32.61	16.42
TubeDETR2022.03		22	43.1	29.7	18.1
STGVT2022.09		21.62	—	29.8	18.94
STGVT2022.03		21.6	—	29.8	18.9
STGRN2022.03		19.8	48.5	25.8	14.6
STGRN2022.09		19.75	48.47	25.77	14.6
VideoMolmo2026.04		15.6	41.7	—	—
WSSTG+L-Net2022.09		14.45	40.86	18	7.89
LLaVA-ST2026.04		14.2	44.1	18.5	7.5
STPR+L-Net2022.09		12.93	40.86	16.27	5.68
GroundeR+L-Net2022.09		11.89	40.86	15.32	5.45
WSSTG+TALL2022.09		11.36	34.63	14.63	5.91
STPR+TALL2022.09		10.4	34.63	12.38	4.27
GroundeR+TALL2022.09		9.78	34.63	11.04	4.09