AI Paper Atlas

Research map / Video / Video generation

VC-VAE: Leveraging Video Codecs for Training-Efficient and High-Fidelity Video VAE

Xinxu Ge, Shang Chai, Litong Gong, Zitong Yu, Xin Liu, Tiezheng Ge

ECCV 2026 · Video › Video generation · approach: Autoregressive long video generation · 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 TDC to causal conv (ablation)Causal Conv Video VAE · Kinetics-600 valPSNR: 33.29 -> 33.64Improved
add TDC to group causal conv (ablation)GCConv Video VAE · Kinetics-600 valPSNR: 33.49 -> 33.79Improved
add TDC at higher temporal compression (ablation)GCConv Video VAE 8x8x8 · Kinetics-600 valPSNR: 31.11 -> 32.01Improved
use in-sequence instead of separate first-frame processing (ablation)VC-VAE · Kinetics-600 valPSNR: 32.15 -> 33.34Improved
initialize from Flux-VAE instead of SD3.5-VAE (no TDC) (ablation)Video VAE w/o TDC · Kinetics-600 valPSNR: 33.13 -> 33.49Improved
add TDC to SD3.5-VAE-initialized model (ablation)Video VAE SD3.5-VAE init · Kinetics-600 valPSNR: 33.13 -> 33.74Improved
add TDC to Flux-VAE-initialized model (ablation)Video VAE Flux-VAE init · Kinetics-600 valPSNR: 33.49 -> 33.79Improved
use VC-VAE over WFVAE at 8x8x8WFVAE · WebVid-10M test (8x8x8)FVD: 44.13 -> 36.63Improved
use VC-VAE over Hunyuan-VAE at 4x8x8 (reconstruction PSNR)Hunyuan-VAE · WebVid-10M test (4x8x8)PSNR: 35.84 -> 36.04Improved
use VC-VAE over Hunyuan-VAE at 4x8x8 (FVD)Hunyuan-VAE · WebVid-10M test (4x8x8)FVD: 36.23 -> 36.80Hurt
use VC-VAE latents for Latte-XL generation instead of IVVAEIVVAE · SkyTimelapsegFVD: 118.6 -> 112.8Improved
use VC-VAE over Hunyuan-VAE on text reconstructionHunyuan-VAE · TokBench-Video 256xT-ACC mean: 61.15 -> 66.24Improved

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