AI Paper Atlas

Research map / Video / Video generation

AUHead: Realistic Emotional Talking Head Generation via Action Units Control

Jiayi Lyu, Leigang Qu, Wenjing Zhang, Hanyu Jiang, Kai Liu, Zhenglin Zhou, Xiaobo Xia, Jian Xue, Tat-Seng Chua

ICLR 2026 · Video › Video generation · approach: Talking face generation · 3 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
generate emotion first then AU (chain of thought) instead of AU only (ablation)Qwen-Audio-Chat · Stage 1 AU predictionMAE: 0.2447 -> 0.2085Improved
generate emotion first then AU (chain of thought) instead of AU only (ablation)Qwen-Audio-Chat · Stage 1 AU predictionF1: 0.62 -> 0.69Improved
predict emotion before AU rather than after AU (ablation)Qwen-Audio-Chat · Stage 1 emotion accuracyACC_emo (%): 51.76 -> 67.01Improved
use 2D landmark or mesh-rendering AU representation instead of 1D AU sequence (ablation)MEMO · MEADFID: 11.1105 -> 10.8701Improved
use 2D landmark or mesh-rendering AU representation instead of 1D AU sequence (ablation)MEMO · CREMAFID: 8.4159 -> 8.3352Improved
use 2D landmark or mesh-rendering AU representation instead of 1D AU sequence (ablation)MEMO · MEADSync: 6.7445 -> 6.6095Hurt
add AU conditioning (AUHead) to MEMOMEMO · MEADFID: 11.1237 -> 10.9671Improved
add AU conditioning (AUHead) to MEMOMEMO · MEADSync: 6.9885 -> 6.6311Hurt
add AU conditioning (AUHead) to HalloV1HalloV1 · MEADSync: 4.9512 -> 6.0201Improved
add AU conditioning (AUHead) to HalloV1HalloV1 · CREMAPSNR: 23.2809 -> 23.0818Hurt
add AU conditioning (AUHead) to HalloV1HalloV1 · MEADPSNR: 22.0258 -> 22.0132No clear effect
raise AU classifier-free guidance scale (ablation)AUHead · MEADFIDMixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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