Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent

• URL: http://arxiv.org/abs/2608.03979v1
• Date: Tue, 04 Aug 2026 17:45:16 GMT
• Title: Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent
• Abstract summary: Video-DR features a decoupled perception-temporal pipeline with stage-wise tool unlocking.<n>Our framework adopts a two-stage training recipe: supervised fine-tuning followed by Group Relative Policy Optimization.<n> Empirical results demonstrate that our Video-ResearchDeepB-A3B establishes a new state-of-the-art of 64.0% average accuracy.
• Score: 81.19460113054048
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We introduce Video-DeepResearch (Video-DR), extending multimodal agents from static images to continuous video streams, a setting that demands dense spatiotemporal grounding coupled with open-web exploration. Preliminary evaluations reveal two critical bottlenecks in current models: (1) modality bias, where agents bypass visual tools in favor of textual search, and (2) parametric knowledge leakage, where models rely on internal memory rather than genuine tool-augmented execution. To address these challenges, we propose Video-DR, featuring a decoupled perception-exploration pipeline with stage-wise tool unlocking that compels exhaustive cross-frame visual grounding prior to web retrieval. Our framework adopts a two-stage training recipe: supervised fine-tuning followed by Group Relative Policy Optimization (GRPO), enabling autonomous exploration that breaks the imitation-learning ceiling. Furthermore, we curate Video-DR-Bench, a human-AI collaborative benchmark comprising 200 complex, multi-hop VQA instances. Empirical results demonstrate that our Video-DeepResearch-35B-A3B establishes a new state-of-the-art of 64.0% average accuracy, surpassing proprietary Claude-4.5-Sonnet (59.0%) by 5.0 points and significantly outperforming GPT-5 (52.5%) and Gemini 2.5 Pro (57.5%). The 30B-A3B variant achieves 59.3%, competitive with Claude-4.5-Sonnet and demonstrating the effectiveness of our training paradigm even at compact scale. Code: https://github.com/Osilly/Vision-DeepResearch.

Related papers

• Kwai Keye-VL-2.0 Technical Report [53.82434681649277]
  Keye-VL-2.0 is a multimodal foundation model designed to advance long-video understanding and agentic intelligence.<n>It is the first to adapt DeepSeek Sparse Attention (DSA) to GQA-based multimodal architectures.<n>Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) paired with Context-RL and Video-RL overcomes the algorithmic dilemma of catastrophic forgetting.
  arXiv  Detail & Related papers  (2026-06-09T09:58:08Z)
• Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding [98.3098451637867]
  Video-MME-v2 is a comprehensive benchmark designed to rigorously evaluate the robustness and faithfulness of video understanding.<n>To guarantee data quality, Video-MME-v2 is constructed through a rigorously controlled human annotation pipeline.
  arXiv  Detail & Related papers  (2026-04-06T17:59:56Z)
• Video-o3: Native Interleaved Clue Seeking for Long Video Multi-Hop Reasoning [28.87800134659646]
  Video-o3 is a novel framework that supports iterative discovery of salient visual clues.<n>Video-o3 substantially outperforms state-of-the-art methods, achieving 72.1% accuracy on MLVU and 46.5% on Video-Holmes.
  arXiv  Detail & Related papers  (2026-01-30T17:47:30Z)
• Video-BrowseComp: Benchmarking Agentic Video Research on Open Web [64.53060049124961]
  Video-BrowseComp is a benchmark comprising 210 questions tailored for open-web agentic video reasoning.<n>It enforces a mandatory dependency on temporal visual evidence, ensuring answers cannot be derived solely through text search.<n>As the first open-web video research benchmark, Video-BrowseComp advances the field beyond passive perception toward proactive video reasoning.
  arXiv  Detail & Related papers  (2025-12-28T19:08:27Z)
• Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval and Synthesis for SLMs [7.3517692707289415]
  We introduce Fathom-DeepResearch, an agentic system composed of two specialized models.<n>The first is Fathom-Search-4B, a DeepSearch model optimized for evidence-based investigation through live web search and targeted webpage querying.<n>The second is Fathom- Synthesizer-4B, trained from Qwen3-4B, which converts multi-turn DeepSearch traces into structured, citation-dense DeepResearch Reports.
  arXiv  Detail & Related papers  (2025-09-28T22:58:11Z)
• VideoScore2: Think before You Score in Generative Video Evaluation [69.43069741467603]
  VideoScore2 is a multi-dimensional, interpretable, and human-aligned framework that explicitly evaluates visual quality, text-to-video alignment, and physical/common-sense consistency.<n>Our model is trained on a large-scale dataset VideoFeedback2 containing 27,168 human-annotated videos.
  arXiv  Detail & Related papers  (2025-09-26T18:09:03Z)
• T*: Re-thinking Temporal Search for Long-Form Video Understanding [66.72243342954823]
  Current temporal search methods only achieve 2.1% temporal F1 score on the Longvideobench subset.<n>Inspired by visual search in images, we propose a lightweight temporal search framework, T* that reframes costly temporal search as spatial search.<n>Extensive experiments show that integrating T* with existing methods significantly improves SOTA long-form video understanding.
  arXiv  Detail & Related papers  (2025-04-03T04:03:10Z)
• CapST: Leveraging Capsule Networks and Temporal Attention for Accurate Model Attribution in Deep-fake Videos [9.209808258321559]
  Attributing a deep-fake to its specific generation model or encoder is vital for forensic analysis, enabling source and tailored countermeasures.<n>We investigate the model attribution problem for deep-fake videos using two datasets: Deepfakes from Different Models (DFDM) and GANGen-Detection.<n>We introduce a novel Capsule-Spatial-Cap (CapST) model that integrates a truncated VGG19 network for feature extraction, capsule networks for temporal extraction.
  arXiv  Detail & Related papers  (2023-11-07T08:05:09Z)

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.