AI Paper Atlas

Research map / Video / Video generation

REGEN: Learning Compact Video Embedding with (Re-)Generative Decoder

Yitian Zhang, Long Mai, Aniruddha Mahapatra, David Bourgin, Yicong Hong, Jonah Casebeer, Feng Liu, Yun Fu

ICCV 2025 · Video › Video generation · approach: Efficient video diffusion models · 2 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 content-aware positional conditioning (vs in-context conditioning) (ablation)REGEN · DAVIS 2019 384x640 unseen resolutionrFVD: 441.98 -> 57.01Improved
add content-aware positional conditioning (vs in-context conditioning) (ablation)REGEN · DAVIS 2019 192x320 training resolutionrFVD: 135.89 -> 128.80Improved
add content-aware positional conditioning (vs in-context conditioning) (ablation)REGEN · DAVIS 2019 384x640 unseen resolutionPSNR: 23.39 -> 29.41Improved
replace MAGVIT-v2 decoder with diffusion decoder at 32x temporal compressionMAGVIT-v2 · DAVIS 2019 256x256rFVD: 1080.15 -> 488.89Improved
replace MAGVIT-v2 decoder with diffusion decoder at 8x temporal compressionMAGVIT-v2 · MCL-JCV 256x256PSNR: 26.61 -> 28.82Improved
replace MAGVIT-v2 with REGEN at base 4x compressionMAGVIT-v2 · MCL-JCV 512x512rFVD: 28.63 -> 22.40Improved
widen MAGVIT-v2 decoder at 16x compression (ablation)MAGVIT-v2 · DAVIS 2019 256x256rFVD: 441.24 -> 486.48Hurt
widen MAGVIT-v2 decoder at 32x compression (ablation)MAGVIT-v2 · DAVIS 2019 256x256rFVD: 1080.15 -> 1020.80Improved
increase diffusion sampling steps (ablation)REGEN 4x · DAVIS 2019 and MCL-JCVPSNR / rFVD: slightly worse PSNR, better rFVDMixed
reduce sampling to one step (ablation)REGEN 4x · reconstruction examples (4x model)visual qualityNo clear effect
add latent extension (SDEdit-style) across chunks (ablation)REGEN · chunk boundary smoothnessx-t slice jumpingImproved
use REGEN decoder latency vs MAGVIT-v2 at 8x8x8MAGVIT-v2 · decoder latency 256x256latency (ms): 317 -> 159Improved

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