Structured Over Scale: Learning Spatial Reasoning from Educational Video

• URL: http://arxiv.org/abs/2601.23251v1
• Date: Fri, 30 Jan 2026 18:20:23 GMT
• Title: Structured Over Scale: Learning Spatial Reasoning from Educational Video
• Abstract summary: We introduce DoraVQA, a dataset of 5,344 question-answer pairs automatically extracted from 8 seasons of Dora the Explorer.<n>We fine-tune both Qwen2 and Qwen3, leveraging the clear correctness signals and structured reasoning traces inherent in educational content.<n>Our approach achieves improvements of 8-14 points on DoraVQA and state-of-the-art 86.16% on CVBench.
• Score: 6.358055069130984
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Vision-language models (VLMs) demonstrate impressive performance on standard video understanding benchmarks yet fail systematically on simple reasoning tasks that preschool children can solve, including counting, spatial reasoning, and compositional understanding. We hypothesize that the pedagogically-structured content of educational videos provides an ideal training signal for improving these capabilities. We introduce DoraVQA, a dataset of 5,344 question-answer pairs automatically extracted from 8 seasons of Dora the Explorer with precise timestamp alignment. Each episode follows a consistent \textit{context-question-pause-answer} structure that creates a self-contained learning environment analogous to interactive tutoring. We fine-tune both Qwen2 and Qwen3 using Group Relative Policy Optimization (GRPO), leveraging the clear correctness signals and structured reasoning traces inherent in educational content. Despite training exclusively on 38 hours of children's educational videos, our approach achieves improvements of 8-14 points on DoraVQA and state-of-the-art 86.16\% on CVBench, with strong transfer to Video-MME and NExT-QA, demonstrating effective generalization from narrow pedagogical content to broad multimodal understanding. Through cross-domain benchmarks, we show that VLMs can perform tasks that require robust reasoning learned from structured educational content, suggesting that content structure matters as much as content scale.

Related papers

• KnowVis: Knowledge-Centric Visual Summarization for Video Lectures [14.336793734778574]
  We propose KnowVis, a framework that transforms linear video lectures into pedagogically grounded visual narratives.<n>KnowVis first extracts a detailed concept map from multimodal video content to identify important and challenging threshold concepts.<n>It then constructs structured knowledge units, and finally synthesizes engaging visual summaries.
  arXiv  Detail & Related papers  (2026-09-03T12:14:34Z)
• Video-IFBench: Evaluating Instruction Following of Multimodal LLMs in Video Understanding Scenarios [61.926801508808616]
  We introduce Video-IFBench, a benchmark for evaluating instruction following in video understanding.<n>We develop an instruction taxonomy with four templates, including single-task, multi-task, selection, and nested instructions.<n>We build a semi-automatic data construction pipeline that combines MLLMs, programmatic processing, and human verification, resulting in 1.5K samples.
  arXiv  Detail & Related papers  (2026-08-26T08:35:42Z)
• Improving Spatial-Temporal Reasoning in Video-Language Models with Structured Video Prompting [0.0]
  Video-language models (VLMs) remain brittle on tasks that require tracking events over time and grounding answers in specific spatial regions.<n>We introduce structured video prompting, a training-free inference-time method that augments the input video with lightweight spatial structure and temporal structure.<n>We evaluate this approach on two complementary video benchmarks and two open video-language models.
  arXiv  Detail & Related papers  (2026-08-23T20:16:09Z)
• EduVQA: Benchmarking AI-Generated Video Quality Assessment for Education [21.304581210099045]
  We present EduAIGV-1k, the first benchmark dataset and evaluation framework dedicated to assessing the quality of AI-generated videos (AIGVs)<n>EduAIGV-1k contains 1,130 short videos produced by ten state-of-the-art text-to-video (T2V) models using 113 pedagogy-oriented prompts.
  arXiv  Detail & Related papers  (2026-03-03T15:05:27Z)
• Less is More: Label-Guided Summarization of Procedural and Instructional Videos [21.13311741987469]
  We propose a three-stage framework, PRISM: Procedural Representation via Integrated Semantic and Multimodal analysis.<n>We analyze adaptive visual sampling, label-driven anchoring, and contextual validation using a large language model (LLM)<n>Our approach generalizes across procedural and domain-specific video tasks, achieving strong performance with both semantic alignment and precision.
  arXiv  Detail & Related papers  (2026-01-18T03:41:48Z)
• Video-LLMs with Temporal Visual Screening [59.18455762289321]
  Temporal Visual Screening (TVS) is a new task that universally pre-processes video question answering and instruction tuning data.<n>TVS is formulated as a modular front-end adapter task that can be seamlessly integrated into both Video Instruction Tuning (training) and Video Question Answering (inference) pipelines.<n> Experiments demonstrate that incorporating TVS yields relative gains of 7.33% (training) and 34.6% (inference)
  arXiv  Detail & Related papers  (2025-08-27T14:33:32Z)
• Video-CoT: A Comprehensive Dataset for Spatiotemporal Understanding of Videos Based on Chain-of-Thought [19.792159494513424]
  Video comprehension is essential for various applications ranging from video analysis to interactive systems.<n>Despite advancements in vision-language models, these models often struggle to capture nuanced,temporal details.<n>To address this, we introduce Video-Thought, a groundbreaking dataset designed to enhance video understanding.
  arXiv  Detail & Related papers  (2025-06-10T14:08:56Z)
• InstructionBench: An Instructional Video Understanding Benchmark [14.71613140347162]
  We introduce InstructionBench, an instructional video understanding benchmark.<n>We formulate Q&A pairs in open-ended and multiple-choice formats to assess both Coarse-Grained event-level and Fine-Grained object-level reasoning.<n>The benchmark finally contains 5k questions across over 700 videos.
  arXiv  Detail & Related papers  (2025-04-07T13:05:09Z)
• STEP: Enhancing Video-LLMs' Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training [87.58996020705258]
  Video Large Language Models (Video-LLMs) have recently shown strong derivation in basic video understanding tasks.<n>Video-LLMs struggle with compositional reasoning that requires multi-step explicit-temporal inference across object relations, interactions and events.<n>We propose STEP, a novel graph-guided self-training method that enables VideoLLMs to generate reasoning-rich finetuning data from any raw videos to improve itself.
  arXiv  Detail & Related papers  (2024-11-29T11:54:55Z)
• E.T. Bench: Towards Open-Ended Event-Level Video-Language Understanding [57.630136434038384]
  We introduce E.T. Bench (Event-Level & Time-Sensitive Video Understanding Benchmark), a large-scale benchmark for open-ended event-level video understanding. We extensively evaluated 8 Image-LLMs and 12 Video-LLMs on our benchmark, and the results reveal that state-of-the-art models for coarse-level (video-level) understanding struggle to solve our fine-grained tasks. Our simple but effective solution demonstrates superior performance in multiple scenarios.
  arXiv  Detail & Related papers  (2024-09-26T17:53:04Z)
• SOK-Bench: A Situated Video Reasoning Benchmark with Aligned Open-World Knowledge [60.76719375410635]
  We propose a new benchmark (SOK-Bench) consisting of 44K questions and 10K situations with instance-level annotations depicted in the videos. The reasoning process is required to understand and apply situated knowledge and general knowledge for problem-solving. We generate associated question-answer pairs and reasoning processes, finally followed by manual reviews for quality assurance.
  arXiv  Detail & Related papers  (2024-05-15T21:55:31Z)
• Learning from Untrimmed Videos: Self-Supervised Video Representation Learning with Hierarchical Consistency [60.756222188023635]
  We propose to learn representations by leveraging more abundant information in unsupervised videos. HiCo can generate stronger representations on untrimmed videos, it also improves the representation quality when applied to trimmed videos.
  arXiv  Detail & Related papers  (2022-04-06T18:04:54Z)
• Video Understanding as Machine Translation [53.59298393079866]
  We tackle a wide variety of downstream video understanding tasks by means of a single unified framework. We report performance gains over the state-of-the-art on several downstream tasks including video classification (EPIC-Kitchens), question answering (TVQA), captioning (TVC, YouCook2, and MSR-VTT)
  arXiv  Detail & Related papers  (2020-06-12T14:07:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.