AI Paper Atlas

Research map / Video / Video-language understanding

Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence

Kun Ouyang, Yuanxin Liu, Linli Yao, Yishuo Cai, Hao Zhou, Fandong Meng, Jie Zhou, Xu Sun

CVPR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add multi-scale (context) frame categorization (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 53.8 -> 57.4Improved
Add evidence difficulty-aware sampling (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 55.2 -> 57.4Improved
Add textual reasoning cold-start stage (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 57.0 -> 57.4Improved
Add multimodal alignment stage (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 56.4 -> 57.4Improved
Add vision-centric reasoning stage (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 53.0 -> 57.4Improved
Add multi-stage cold-start before RLVR (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 51.0 -> 57.4Improved
Add identification reward (ablation)AIR RLVR · six video reasoning benchmarks (avg)accuracy: 53.8 -> 57.4Improved
Add retrieval reward (ablation)AIR RLVR · six video reasoning benchmarks (avg)accuracy: 54.0 -> 57.4Improved
Use Conan-style evidence retrieval instead of single-round text CoT (ablation)Conan · six video reasoning benchmarks (avg)accuracy: 55.2 -> 57.4Improved
Skip textual reasoning cold-start stage (ablation)Conan · VRBenchaccuracy: 81.4 vs 81.0No clear effect
Train Conan from Qwen2.5-VL-7B-InstructQwen2.5-VL-7B-Instruct · Video-Holmesaccuracy: 28.5 -> 44.6Improved
Train Conan from Qwen2.5-VL-7B-InstructQwen2.5-VL-7B-Instruct · long video understanding (avg of four)accuracy: 48.0 -> 54.9Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy