# jixiaozhong/Sonic

Official implementation of "Sonic: Shifting Focus to Global Audio Perception in Portrait Animation"

Repository: https://github.com/jixiaozhong/Sonic
Canonical: https://ross.abutalabs.com/products/jixiaozhong-sonic
Language: Python
License: NOASSERTION
License Family: other
Last push: 2026-01-08T11:32:14+00:00

## Health v2 (maintenance only)
Score: 49/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 61, release rhythm 35, longevity 46
- inputs: {"age_days": 645, "days_push": 237, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3273, forks 291 (observed 2026-08-28T04:07:53.212056+00:00)

## What it is
Sonic is the official PyTorch implementation of the CVPR 2025 paper 'Sonic: Shifting Focus to Global Audio Perception in Portrait Animation'. It animates a still portrait image using an audio clip to generate a talking-head video with natural motion and lip sync.

## Use cases
- animate a portrait photo with audio to make a talking head video
- generate lip-synced avatar videos from a single image and voice clip
- create talking portrait animations for research on audio-driven video generation
- bring anime or sculpture images to life with speech audio
- reproduce CVPR 2025 portrait animation research results

## When to choose
- you need audio-driven portrait animation from a single image
- you want a research-grade talking-head model with a paper backing
- you need non-photorealistic support like anime portraits
- you want to try it quickly via the Gradio or Hugging Face demo

## When to avoid
- you need a commercially licensed solution - the license is CC BY-NC-SA (non-commercial)
- you lack a GPU or cannot run large diffusion models locally
- you need real-time streaming avatar generation
- you want a polished end-user app rather than a research codebase

## Facets
- artifact type: library
- maturity: active
- function: video-processing, audio-processing, machine-learning, deep-learning, image-processing
- domain: computer-vision, artificial-intelligence, deep-learning
- platform: python, cross-platform
- tags: portrait-animation, talking-head, audio-driven, lip-sync, cvpr-2025, diffusion, non-commercial-license, gradio-demo, video, audio, gpu, linux

## Member repositories
- jixiaozhong/Sonic (main) score 49

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:53.212056+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:43:48.432857+00:00, confidence not recorded.
  - readme: https://github.com/jixiaozhong/Sonic (fetched 2026-08-28T04:07:53.212056+00:00, sha 082089d010e2)
- Data as of 2026-08-30T08:39:29.467469+00:00.