AI Paper Atlas

Research map / Video / Video-language understanding

VidGuard-R1: AI-Generated Video Detection and Explanation via Reasoning MLLMs and RL

Kyoungjun Park, Yifan Yang, Juheon Yi, Shicheng Zheng, Muhammad Muaz, Yifei Shen, Dongqi Han, Caihua Shan, Lili Qiu

ICLR 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 13 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 temporal-artifact reward (GRPO-TA) to GRPOVidGuard-R1 (GRPO) · Own dataset (CogVideoX)accuracy (%): 81.30 -> 82.17Improved
add diffusion-step quality reward (GRPO-Q) to GRPOVidGuard-R1 (GRPO) · Own dataset (CogVideoX)accuracy (%): 81.30 -> 84.32Improved
add diffusion-step quality reward (GRPO-Q) to GRPOVidGuard-R1 (GRPO) · Own dataset (HunyuanVideo)accuracy (%): 81.90 -> 86.17Improved
replace DPO with GRPOVidGuard-R1 · Own dataset (CogVideoX)accuracy (%): 79.13 -> 81.30Improved
add RL (DPO) after SFT cold startVidGuard-R1 (CoT) · Own dataset (CogVideoX)accuracy (%): 66.18 -> 79.13Improved
increase intermediate diffusion steps from 1 to 5 (ablation)VidGuard-R1 (GRPO-Q) · Own datasetaccuracy (%): 81.63 -> 85.80Improved
increase diffusion steps from 1 to 3 (ablation)VidGuard-R1 (GRPO-Q) · Own datasetaccuracy (%): 81.63 -> 83.21Improved
set alpha1=0.5, alpha2=0.3 versus 0.3/0.1 (ablation)VidGuard-R1 (GRPO-TA) · Own datasetaccuracy (%): 81.31 -> 83.57Improved
raise reward weights to alpha1=0.7, alpha2=0.5 from 0.5/0.3 (ablation)VidGuard-R1 (GRPO-TA) · Own datasetaccuracy (%): 83.57 -> 82.53Hurt
add GenVideo data to own-dataset training (ablation)VidGuard-R1 (GRPO) · Own datasetaccuracy (%): 81.65 -> 82.97Improved
add own data to GenVideo training (ablation)VidGuard-R1 (GRPO) · GenVideoaccuracy (%): 97.53 -> 97.98Improved
pretrain on GenVideo then zero-shot to GenVidBench vs training on itVidGuard-R1 (GRPO) · GenVidBenchmean Top-1 accuracy (%): 97.53 -> 96.37Mixed

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