AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Sep-Stereo: Visually Guided Stereophonic Audio Generation by Associating Source Separation

Hang Zhou, Xudong Xu, Dahua Lin, Xiaogang Wang, Ziwei Liu

ECCV 2020 · Vision-language and multimodal › Cross-modal learning · approach: Audio-visual source localization · 79 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 Associative-Conv visual-audio fusion to Mono2Binaural (ablation)Mono2Binaural · FAIR-PlaySTFT distance: 0.959 -> 0.893Improved
add Associative-Conv visual-audio fusion to Mono2Binaural (ablation)Mono2Binaural · YT-MusicSTFT distance: 1.346 -> 1.147Improved
expand Associative-Conv to multi-scale APNet (ablation)Sep-Stereo · YT-MusicSTFT distance: 1.147 -> 1.070Improved
expand Associative-Conv to multi-scale APNet (ablation)Sep-Stereo · FAIR-PlaySTFT distance: 0.893 -> 0.889Improved
add separative learning with mono MUSIC data to APNet (ablation)Sep-Stereo · FAIR-PlaySTFT distance: 0.889 -> 0.879Improved
add separative learning with mono MUSIC data to APNet (ablation)Sep-Stereo · YT-MusicSTFT distance: 1.070 -> 1.051Improved
add separative learning to Mono2Binaural via feature rearrangement (ablation)Mono2Binaural · FAIR-PlaySTFT distance: 0.959 -> 0.930Improved
add output-difference loss L_D (ablation)Sep-Stereo · FAIR-PlaySTFT distance: 0.976 -> 0.879Improved
use APNet left/right branch outputs instead of direct U-Net difference output (ablation)Sep-Stereo · FAIR-PlaySTFT distance: 0.933 -> 0.879Improved
use horizontal placement instead of vertical extremes in feature rearrangement (ablation)Sep-Stereo · FAIR-PlaySTFT distance: 0.919 -> 0.879Improved
add APNet-based separation with shared backbone versus Associative-Conv (ablation)Sep-Stereo · MUSIC source separationSDR: 5.79 -> 8.07Improved
Sep-Stereo versus PixelPlayer for separationSep-Stereo · MUSIC source separationSDR/SIR/SAR: SDR 7.67 -> 8.07, SIR 14.81 -> 10.14, SAR 11.24 -> 15.51Mixed

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