AI Paper Atlas

Research map / Video / Video generation

MotionEnhancer: Leveraging Video Diffusion for Motion-Enhanced Vision-Language Models

Yifan Xu, Chao Zhang, Ruifei Ma, Fei Gao, Zhifei Yang, Jiaxing Qi, Zhipeng Chen

CVPR 2026 · Video › Video generation · approach: Efficient video diffusion models · 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 Motion-sensitive Head Selection (MHS) (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 54.83 -> 56.60Improved
add Motion-sensitive Head Selection (MHS) (ablation)MotionEnhancer (Qwen2.5-VL-7B) · FAVOR-Bench close-endedOverall accuracy: 44.83 -> 46.65Improved
add Motion-salient Text Token Identification (MTTI) (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 54.83 -> 55.80Improved
add MTTI on top of MHS (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 56.60 -> 57.04Improved
add MHS on top of MTTI (ablation)MotionEnhancer (Qwen2.5-VL-7B) · FAVOR-Bench close-endedOverall accuracy: 45.47 -> 46.88Improved
use 5 DDIM steps instead of 2 (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 52.76 -> 57.04Improved
use 5 DDIM steps instead of 2 (ablation)MotionEnhancer (Qwen2.5-VL-7B) · FAVOR-Bench close-endedOverall accuracy: 24.01 -> 46.88Improved
use 10 DDIM steps instead of 5 (ablation)MotionEnhancer (Qwen2.5-VL-7B) · FAVOR-Bench close-endedOverall accuracy: 46.88 -> 49.02Improved
use 10 DDIM steps instead of 5 (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 57.04 -> 57.51Improved
select 50% heads/tokens instead of 25% (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 55.33 -> 57.04Improved
select 50% heads/tokens instead of 25% (ablation)MotionEnhancer (Qwen2.5-VL-7B) · FAVOR-Bench close-endedOverall accuracy: 47.12 -> 46.88Hurt
select 50% heads/tokens instead of 75% (ablation)MotionEnhancer (Qwen2.5-VL-7B) · MotionBench devOverall accuracy: 54.88 -> 57.04Improved

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