- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper proposes anchorPlay, which uses one video model as both questioner and localizer for self-playing training. A frozen version of the reference model proposes intervals and boxes, which are checked across different views of the video. The accepted targets are used for reliability-weighted RL and a coordinate prediction loss. Questions change with the solver's ability. The paper reports temporal and spatiotemporal grounding results, plus transfer to other visual tasks.
Critical strengths
- This paper identifies the challenge of temporal grounding annotations. It proposes self-play training to avoid additional human-provided questions or localization annotations.
- It uses a frozen reference to keep the localization targets separate from the solver’s current predictions. The questions can still adapt to the solver’s ability, without using its own answers to define the targets.
Critical weaknesses
- The reference can be consistently wrong. The same frozen model proposes the target, checks it across changed sampling offsets, sizes and reflections, judges query validity, and supplies the coordinate targets. An early start, late end, or wrong subject may survive those checks and become supervision.
- Missing comparison with VideoSSR (CVPR 2026). It uses synthetic video tasks with known targets, including anomaly intervals, and reports temporal-grounding transfer with Qwen3-VL-8B. It is cited in Section 2 but not compared in the experiments.
- The anchor design seems only useful for single segment. What if the question needs to localize multiple segments? This method seems hard to generalize to multi-segment localizations.
- Some numbers are in consistent. Why in Table 8 the validity of Qwen3.5-9B on Charades is only 5.6%, but in Table 1 it can still achieve mIoU of 59.26.
- The training cost is unclear. The method adds anchor generation and checks across several reference views, then generates eight questions and eight answers for each valid question. The SFT + AnchorPlay model remains below the full OneThinker reference on temporal grounding. That could be a useful tradeoff if it saves enough annotation or compute, but the paper does not quantify this.
Questions for authors
- What will be the anchorplay result if initialized from OneThinker-8B.
- There is a dagger sign in Table 1 behind OneThinker-8B, what is the meaning of the dagger? There is no explanation for that.
- It is better to clearly write the name of the base model in the Table to avoid confusion.
- It seems that the anchorplay result is still far from the full annotated OneThinker-8B. One might question about the effectiveness. What if the method starts from a strong baseline like OneThinker-8B, would OneThinker-8B boost the performance?
- The anchor design seems only useful for single segment. What if the question needs to localize multiple segments? Would this method be applied to multi-segment localization? QVHighlights contains multi-segment localization questions, how would the method perform on this benchmark.
- It is quite interesting that the pure temporal grounding training can also enhance image grounding. But it is a little bit overclaimed, since the training data has SVTG task, which still contains spatial grounding.
- Table 5 needs checking. It says all ablations use 40 training steps, but the full model's scores (39.27 TG / 30.08 STVG) match step 60 in Table 7. The step-40 scores are about 37.64 / 27.07. I cannot tell whether the full model and the ablations were trained for the same number of steps.