AI Paper Atlas

Research map / Video / Video-language understanding

Less Is More: Vision Representation Compression for Efficient Video Generation with Large Language Models

Yucheng Zhou, Jihai Zhang, Guanjie Chen, Jianbing Shen, Yu Cheng

AAAI 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 3 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 representation compression (ratio 4) to Loong-style LLM video generatorLoong · Class-to-video FaceForensicsFVD: 46.11 -> 26.64Improved
add representation compression to LoongLoong · Class-to-video SkyTimelapseFVD: 62.71 -> 41.95Improved
add representation compression to LoongLoong · Class-to-video UCF101FVD: 254.47 -> 250.53Improved
add representation compression to LoongLoong · Class-to-video Taichi-HDFVD: 105.53 -> 96.39Improved
compress tokens 4x instead of full sequenceLoong · 17-frame generation time, Base 100Minference time (s): 12.32 -> 0.88Improved
compress tokens 4x instead of full sequenceLoong · 17-frame generation time, XLarge 770Minference time (s): 26.08 -> 2.89Improved
raise compression rate from 4 to 16 or 64 (ablation)VRC · Class-to-video datasetsFVDHurt
moderate compression (rate 4) versus no compression (ablation)VRC · Class-to-video datasetsFVDImproved
scale model from 100M to 770M (ablation)VRC · Class-to-video datasetsFVDImproved
use VRC vs Loong for memoryVRC · GPU memorymemoryImproved
VRC vs VideoPoet on zero-shot text-to-videoVRC · MSR-VTTFVD: 213 -> 268Hurt
VRC (1.2B) vs Loong (7B) on zero-shot text-to-videoVRC · MSR-VTTCLIPSIM: 0.2903 -> 0.2956Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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