AI Paper Atlas

Research map / Video / Video generation

EMO: Emote Portrait Alive - Generating Expressive Portrait Videos with Audio2Video Diffusion Model under Weak Conditions

Linrui Tian, Qi Wang, Bang Zhang, Liefeng Bo

ECCV 2024 · Video › Video generation · approach: Talking face generation · 214 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 speed layers with an assigned speed value (ablation)EMO · HDTFVelocity Variance: 1.002 -> 0.357Improved
add speed layers (ablation)EMO · HDTFVMV: 0.134 -> 0.046Improved
raise assigned speed value from 0.1 to 1.9 (ablation)EMO speed layers · HDTFMean Velocity: 0.878 -> 1.246Improved
add 250 hours of collected training data (ablation)EMO · HDTFE-FID: 0.215 -> 0.116Improved
add 250 hours of collected training data (ablation)EMO · HDTFFVD: 102.78 -> 67.66Improved
use direct audio-to-video diffusion instead of 3DMM or landmark intermediatestalking head generation · HDTFE-FID: 0.503 -> 0.116 (vs SadTalker)Improved
EMO versus Wav2Lip on lip synctalking head generation · HDTFSyncNet: 5.79 -> 3.89Hurt
train speed layers jointly with audio layers (ablation)EMO · training strategyaudio driving abilityHurt

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