Spatio-Temporal Video Grounding on VidSTG Declarative (test)
34m_vIoUCG-STVG
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| CG-STVGLearning Paradigm=Fully-Supervised2026.02 | 34 | 51.4 | 47.7 | 33.1 | |
| CSDVLLearning Paradigm=Fully-Supervised2026.02 | 33.7 | — | 47.2 | 32.8 | |
| STCATLearning Paradigm=Fully-Supervised2026.02 | 33.1 | 50.8 | 46.2 | 32.6 | |
| TubeDETRLearning Paradigm=Fully-Supervised2026.02 | 30.4 | 48.1 | 42.5 | 28.2 | |
| ASTGLearning Paradigm=Zero-Shot2026.02 | 29.2 | 45.6 | 40.3 | 27.8 | |
| RealVGLearning Paradigm=Zero-Shot2026.02 | 29 | 37.4 | 35 | 25.5 | |
| ReCLIPLearning Paradigm=Zero-Shot2026.02 | 19.1 | — | 29.4 | 10.6 | |
| LN-STVGLearning Paradigm=Zero-Shot2026.02 | 18 | — | 29.8 | 12.2 | |
| E3MLearning Paradigm=Zero-Shot2026.02 | 16.2 | — | 20.5 | 11.9 | |
| CoSPaLLearning Paradigm=Weakly-Supervised2026.02 | 16 | 41.1 | 20.1 | 13.1 | |
| STProLearning Paradigm=Weakly-Supervised2026.02 | 15.5 | — | 19.4 | 12.7 | |
| VCMALearning Paradigm=Weakly-Supervised2026.02 | 14.5 | — | 18.6 | 8.8 | |
| WINNERLearning Paradigm=Weakly-Supervised2026.02 | 11.6 | — | 14.1 | 7.4 | |
| RedCircleLearning Paradigm=Zero-Shot2026.02 | 8.6 | — | 7.6 | 0.9 |