VideoKR: Towards Knowledge- and Reasoning-Intensive Video Understanding
Quick Answer
VideoKR introduces a large-scale corpus with 315K video reasoning examples from 145K expert-domain videos, enhancing knowledge-intensive video understanding.
Quick Take
VideoKR introduces a large-scale corpus with 315K video reasoning examples from 145K expert-domain videos, enhancing knowledge-intensive video understanding. Models trained on VideoKR outperform previous methods in knowledge-intensive reasoning while maintaining competitive performance in general video reasoning, emphasizing the importance of data design.
Key Points
- VideoKR consists of 315K reasoning examples over 145K expert-domain videos.
- The corpus supports deeper video reasoning capabilities with diverse and reliable examples.
- VideoKR-Eval is a new benchmark requiring genuine video understanding.
- Models post-trained on VideoKR show superior performance in knowledge-intensive reasoning.
- Comprehensive ablations provide insights for future advancements in video reasoning.
Article Excerpt
From source RSS / original summaryarXiv:2606. 05259v1 Announce Type: new Abstract: We introduce VideoKR, the first large-scale training corpus specifically designed to strengthen knowledge- and reasoning-intensive video understanding. It comprises 315K video reasoning examples over 145K newly collected, CC-licensed, expert-domain videos.
We develop a human-in-the-loop, skill-oriented example generation pipeline that targets progressively deeper video reasoning capabilities while ensuring the difficulty, diversity, and reliability of both the examples and their CoT rationales. We also curate VideoKR-Eval, a new expert-annotated benchmark where questions require genuine video understanding and knowledge-intensive reasoning rather than textual shortcuts.
Our experiments show that, under a standard SFT$\rightarrow$GRPO pipeline, models post-trained on VideoKR outperform prior post-training approaches on knowledge-intensive video reasoning while remaining competitive on general video reasoning, highlighting data design as a key driver of progress in video reasoning. We further conduct comprehensive ablations to isolate the contributions of VideoKR, providing actionable insights for future work.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.