# MRzzm/DINet

The source code of "DINet: deformation inpainting network for realistic face visually dubbing on high resolution video."

Repository: https://github.com/MRzzm/DINet
Canonical: https://ross.abutalabs.com/products/dinet
Language: Python
License Family: other
Last push: 2023-09-25T23:06:26+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 98
- inputs: {"age_days": 1373, "days_push": 1073, "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 1128, forks 181 (observed 2026-08-28T04:03:41.990411+00:00)

## What it is
DINet is the official PyTorch implementation of an AAAI 2023 paper on realistic face visually dubbing, which deforms and inpaints mouth regions in high-resolution video to match driving audio. It includes inference scripts with pretrained models and a coarse-to-fine training pipeline for arbitrary resolutions.

## Use cases
- dub a talking-head video to match new audio
- generate lip-synced face video from speech
- replace mouth region in high resolution video driven by audio
- train an audio-driven face dubbing model on custom dataset
- process HDTF dataset videos for talking face generation

## When to choose
- you need research-grade audio-driven face dubbing with high resolution output
- you want to retrain or fine-tune a talking-face model on your own audio-visual data
- your videos are frontal-view talking heads with normal lighting

## When to avoid
- you need a production-ready tool with strong generalization to arbitrary faces and angles
- you need a no-GPU or plug-and-play consumer application
- you require a maintained project with a license or active development

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, video-processing, image-processing, audio-processing, machine-learning
- domain: computer-vision, deep-learning, artificial-intelligence
- platform: python, windows
- tags: talking-face, face-dubbing, lip-sync, video-inpainting, research-code, aaai2023, audio-driven, video, audio, linux, gpu

## Member repositories
- MRzzm/DINet (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:41.990411+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 2026-08-30T06:38:28.600599+00:00, confidence not recorded.
  - readme: https://github.com/MRzzm/DINet (fetched 2026-08-28T04:03:41.990411+00:00, sha 8ed0fbd0c8ef)
- Data as of 2026-08-30T08:39:29.467469+00:00.