Skip to content

Source paper

PersonaLive! Expressive Portrait Image Animation for Live Streaming

Zhiyuan Li, Chi-Man Pun, Chen Fang and 2 more

Zhiyuan Li, Chi-Man Pun, Chen Fang, Jue Wang, Xiaodong Cun

arXiv 2512.11253 · 202527 upvotes on Hugging Face

Research ideas from this paper (4)

Generated by AI. The paper’s authors did not write these.

Temporal Redundancy Exploitation via Feature-Space Propagation

A system that integrates motion-aware feature propagation into latent diffusion models to skip redundant denoising steps for stable background regions, significantly increasing video generation frame rates.

PersonaLive! Expressive Portrait Image Animation for Live Streaming
🔴Hard💡Moderate InnovationSingle GPU or lessNo close match found#Generative media#Efficiency

Cross-Domain Adaptation for Virtual and Non-Human Avatars

A framework that introduces a domain-agnostic motion retargeting layer using a contrastive appearance prior and parameter-efficient fine-tuning, allowing human portrait generators to generalize to cartoons and animals.

PersonaLive! Expressive Portrait Image Animation for Live Streaming
🟠Medium💡Moderate InnovationSingle GPU or lessNo close match found#Generative media#Vision

Reference-Conditioned VAE for High-Fidelity Streaming

A reference-conditioned variational autoencoder that utilizes feature skip-connections from the initial source portrait into the decoder stages, combining lightweight streaming speed with high-frequency visual fidelity.

PersonaLive! Expressive Portrait Image Animation for Live Streaming
🟠Medium💡Moderate InnovationSingle GPU or lessNo close match found+2#Generative media#Efficiency

Semantic-Aware Historical Keyframe Selection

A learned keyframe selection mechanism that predicts the stability value of generated frames using motion similarity and identity confidence scores, preventing long-term identity drift in streaming generation.

PersonaLive! Expressive Portrait Image Animation for Live Streaming
🟡Fairly easy💡Moderate InnovationSingle GPU or lessNo close match found#Generative media#Vision

AI summary

PersonaLive is a diffusion-based framework designed for real-time, expressive portrait image animation specifically optimized for live streaming. The paper addresses the high latency of traditional diffusion models by combining few-step distillation with an autoregressive micro-chunk generation paradigm to achieve up to 22x speedup while maintaining temporal stability.

How it works (AI summary)

The architecture uses a three-stage pipeline: (1) hybrid motion control combining 1D facial motion embeddings and 3D implicit keypoints for global/local dynamics; (2) a fewer-step appearance distillation strategy using adversarial and LPIPS losses to compress denoising into 1-4 steps; and (3) an autoregressive micro-chunk streaming paradigm with a Historical Keyframe Mechanism (HKM) to mitigate long-term error accumulation.