AI Paper Atlas

Research map / Video / Video segmentation and tracking

Audio-Assisted Face Video Restoration with Temporal and Identity Complementary Learning

Yuqin Cao, Yixuan Gao, Wei Sun, Xiaohong Liu, Yulun Zhang, Xiongkuo Min

AAAI 2026 · Video › Video segmentation and tracking · approach: Deepfake video detection · 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
add identity features (landmark + audio intra-frame module) (ablation)GAVN · VoxCeleb2 compression artifact removalPSNR: 28.7797 -> 28.9780Improved
add identity features (ablation)GAVN · VoxCeleb2 deblurPSNR: 38.8352 -> 39.3441Improved
add identity features (ablation)GAVN · VoxCeleb2 super-resolutionPSNR: 35.7568 -> 36.1462Improved
add audio features (ablation)GAVN · VoxCeleb2 compression artifact removalPSNR: 28.9480 -> 28.9780Improved
add audio features (ablation)GAVN · VoxCeleb2 deblurPSNR: 39.0181 -> 39.3441Improved
add audio features (ablation)GAVN · VoxCeleb2 super-resolutionPSNR: 35.8427 -> 36.1462Improved
add audio features (ablation)GAVN · VoxCeleb2 compression artifact removalLPIPS: 0.1718 -> 0.1658Improved
GAVN vs audio-aided DAVD-NetGAVN · VoxCeleb2 compression artifact removalPSNR: 28.7269 -> 28.9780Improved
GAVN vs audio-aided DAVD-NetGAVN · Obama compression artifact removalPSNR: 28.6645 -> 30.0221Improved
GAVN vs BasicVSR++ at low resolution scale 2GAVN · VoxCeleb2 super-resolution distortion levelPSNR: 42.4803 -> 44.1499Improved
stronger compression (CRF 35 vs 45) gain of GAVN over DAVD-NetGAVN · VoxCeleb2 compressionPSNR: +0.1130 at CRF 35; +0.2923 at CRF 45Mixed
GAVN on real-world degraded videos vs DAVD-NetGAVN · Real-world videosSyncNet confidence: 1.6581 -> 1.7895Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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