Metraipolis › archive › dc_c3cc98cf2256

Counterfactual Attention Policy Distillation for Temporal Video Grounding — arXiv:2609.34581

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34581
License (as archived)
CC0-1.0
Archived
2026-10-07T23:52:17.071Z
SHA-256 of the archived text
8827b6203cef986bddadd19f095e772123f263303f6c479f186ee9062fa29ae5
Stable URL of this page
https://metraipolis.com/documento/dc_c3cc98cf2256

How to cite

Counterfactual Attention Policy Distillation for Temporal Video Grounding — arXiv:2609.34581 — arXiv (en), https://arxiv.org/abs/2609.34581 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c3cc98cf2256 (Metraipolis archive dc_c3cc98cf2256, sha256 8827b6203cef986bddadd19f095e772123f263303f6c479f186ee9062fa29ae5)
BibTeX
@misc{dcc3cc98cf2256,
  title = {Counterfactual Attention Policy Distillation for Temporal Video Grounding — arXiv:2609.34581},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34581},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c3cc98cf2256; sha256 8827b6203cef986bddadd19f095e772123f263303f6c479f186ee9062fa29ae5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34581. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Temporal video grounding is a key capability of advanced Multimodal Large Language Models (MLLMs) for the thorough understanding of video events, which is however often limited by repeated actions and visually similar contexts in long videos. In this paper, we study this issue from the perspective of On-policy distillation (OPD) and propose a new training regime for MLLMs termed Counterfactual Attention Policy Distillation (CAPD). In particular, OPD is a viable solution for MLLMs via providing dense teacher supervision on student-generated trajectories. But its next-token based teacher-student distillation is hard to identify the specific video segments supporting each predicted timestamp, which is critical for temporal grounding. In this case, CAPD measures how masking each temporal group changes the teacher's output distribution. The resulting counterfactual influence calibrates the teacher's attention and weights token-level distillation, allowing the student to learn the temporal evidence that affects boundary prediction. To validate CAPD, we trained it on Qwen3-VL-8B-Instruct using only 2,500 samples for one epoch, and evaluated it on the TimeLens and multiple general video benchmarks. Experimental results show that CAPD improves average recall by 12.0% relative to GRPO on TimeLens while preserving general video understanding, achieving comparable accuracy to the base model.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.34581 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T23:29:27.402Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).