AI Paper Atlas

Research map / Video / Video-language understanding

VideoVeritas: AI-Generated Video Detection via Perception Pretext Reinforcement Learning

Hao Tan, jun lan, Senyuan Shi, Zichang Tan, Zijian Yu, Huijia Zhu, Weiqiang Wang, Jun Wan, Zhen Lei

ICML 2026 · Video › Video-language understanding · approach: Long video understanding · 0 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 SFT (4K samples) to base model (ablation)Qwen3-VL-8B · MintVid Othersaccuracy: 55.2 -> 68.3Improved
add SFT (ablation)Qwen3-VL-8B · MintVid Factaccuracy: 91.9 -> 90.7Hurt
add joint DPO after SFT (ablation)SFT model · MintVid Factaccuracy: 90.7 -> 94.6Improved
add response-level DPO_t to J-DPO (ablation)J-DPO · MintVid Othersaccuracy: 66.3 -> 69.1Improved
add video-level DPO_v to J-DPO (ablation)J-DPO · MintVid Factaccuracy: 92.9 -> 94.6Improved
add J-DPO cold start before RL (vs pure RL) (ablation)VideoVeritas · MintVid Othersaccuracy: 69.1 -> 79.2Improved
add format reward to pure RL (ablation)Pure RL · MintVid Factaccuracy: 96.1 -> 87.5Hurt
add SSL object-counting perception task to PPRL (ablation)VideoVeritas · MintVid Phantom-14Baccuracy: 77.1 -> 81.0Improved
add SSL + G-G perception tasks (ablation)VideoVeritas · MintVid Seedance1.0Proaccuracy: 81.3 -> 86.6Improved
add artifact grounding (A-G) on top of SSL + G-G (ablation)VideoVeritas · MintVid Jimeng3.0Proaccuracy: 80.8 -> 74.4Hurt
add 1K perception data (ablation)PPRL · MintVidaccuracy: +1.1% Acc and 1.3% F1Improved
scale perception data beyond ~5K (ablation)PPRL · MintVidaccuracyNo clear effect

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