AI Paper Atlas

Research map / Video / Video generation

AVTok: 1D Unified Tokenization for Holistic Audio-Video Generation

Trung Kien Pham, I -Chieh Chen, Qifeng Chen, Long Chen

ECCV 2026 · Video › Video generation · approach: Video-to-audio synthesis · 1 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
use dual-stream design with modality-specific queries and norms instead of single-stream (ablation)AVTok · VGGSound reconstructionrFAD: 10.26 -> 5.93Improved
use dual-stream design instead of single-stream (ablation)AVTok · VGGSound reconstructionrFVD: 14.87 -> 12.80Improved
add VFAL hierarchical training (ablation)AVTok · VGGSound cJAVGgFVD: 193.28 -> 138.80Improved
add VFAL hierarchical training (ablation)AVTok · VGGSound reconstructionrFAD: 9.38 -> 5.93Improved
add representation alignment loss L_rep (ablation)AVTok · VGGSound cJAVGgFAD: 75.09 -> 56.58Improved
add representation alignment loss L_rep (ablation)AVTok · VGGSound reconstructionrFAD: 8.48 -> 5.93Improved
add cross-modal AR prior loss L_prior (ablation)AVTok · VGGSound cJAVGgFVD: 249.47 -> 138.80Improved
add cross-modal AR prior loss L_prior (ablation)AVTok · VGGSound reconstructionrFVD: 10.63 -> 12.80Hurt
replace CAV-MAE with CAV-MAE Sync as alignment model (ablation)AVTok · VGGSound reconstructionrFAD: 6.40 -> 5.93Improved
replace HiFi-GAN with BigVGAN vocoder (ablation)AVTok · VGGSound reconstructionrFAD: 5.93 -> 5.72Improved
halve video tokens from 1024 to 512 (ablation)AVTok · VGGSound reconstructionrFAD: 5.93 -> 14.90Hurt
halve audio tokens from 128 to 64 (ablation)AVTok · VGGSound reconstructionrFAD: 5.93 -> 12.77Hurt

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