Metraipolis › archive › dc_450f102afa46
FinishStrong: Learning to End Well after Imperfect Start in Video Temporal Grounding — arXiv:2611.01765
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.01765
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:41:39.577Z
- SHA-256 of the archived text
0efeb25a6240321700db2282d03fdf19f34886585665523b6ba4d6f4c3424885- Stable URL of this page
https://metraipolis.com/documento/dc_450f102afa46
How to cite
FinishStrong: Learning to End Well after Imperfect Start in Video Temporal Grounding — arXiv:2611.01765 — arXiv (en), https://arxiv.org/abs/2611.01765 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_450f102afa46 (Metraipolis archive dc_450f102afa46, sha256 0efeb25a6240321700db2282d03fdf19f34886585665523b6ba4d6f4c3424885)
BibTeX
@misc{dc450f102afa46,
title = {FinishStrong: Learning to End Well after Imperfect Start in Video Temporal Grounding — arXiv:2611.01765},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.01765},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_450f102afa46; sha256 0efeb25a6240321700db2282d03fdf19f34886585665523b6ba4d6f4c3424885},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.01765. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
A poor start lowers the ceiling, but does not negate the value of finishing well. In autoregressive video temporal grounding, models systematically exhibit an endpoint performance asymmetry, predicting ends significantly worse than starts. A committed flawed start acts as an immutable physical constraint, irreversibly capping the attainable overlap as a sunk loss. While models should minimize the remaining conditional regret, standard post-training paradigms fail to teach this adaptation. Specifically, teacher-forced SFT suffers from strict prefix exposure bias, and holistic trajectory evaluation in standard GRPO inadvertently penalizes conditionally optimal ends if they follow a poor start. To address this, we introduce FinishStrong, a novel two-stage post-training framework explicitly designed to teach optimal continuations. First, Online-start SFT exposes the end prediction to the model's own generated contexts. Second, Conditional-tree GRPO restructures the sampling topology, centering end rewards exclusively within the same-start branch to mathematically isolate the conditional end quality from the inherited start penalty. FinishStrong-9B achieves a state-of-the-art 61.7\% overall IoU on TimeLens-Bench, remarkably outperforming flagship closed-source models like Gemini-3.8-Flash and Qwen3.8-Max. Crucially, when forced to continue from imperfect starts, it reduces the subsequent end prediction error by 36.5\%, proving its exceptional ability to end well after flawed starts. Code will be open. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2611.01765 — 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 2026-11-03T11:06:37.511Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).