# ant-research/CoDeF

[CVPR'24 Highlight] Official PyTorch implementation of CoDeF: Content Deformation Fields for Temporally Consistent Video Processing

Repository: https://github.com/ant-research/CoDeF
Canonical: https://ross.abutalabs.com/products/codef
Homepage: https://qiuyu96.github.io/CoDeF/
Language: Python
License: NOASSERTION
License Family: other
Topics: editing, video
Last push: 2024-04-07T03:27:29+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 79
- inputs: {"age_days": 1114, "days_push": 878, "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 4846, forks 382 (observed 2026-08-28T04:09:01.344488+00:00)

## What it is
CoDeF is the official PyTorch implementation of Content Deformation Fields, a video representation combining a canonical content field and a temporal deformation field. It lifts image algorithms like image-to-image translation and keypoint detection to temporally consistent video processing without training.

## Use cases
- apply image editing styles consistently across a whole video
- convert image-to-image translation models into video-to-video translation
- track keypoints in video without training a tracker
- track non-rigid objects like water and smoke
- research on neural video representations

## When to choose
- you need temporally consistent video editing or translation from a single-image algorithm
- you want training-free keypoint tracking in videos
- you have a CUDA GPU and want to reproduce CVPR 2024 research results

## When to avoid
- you need a production-ready video editing tool with a GUI
- you lack a CUDA GPU (10GB+ VRAM recommended)
- you need real-time video processing, since per-video field optimization is slow

## Facets
- artifact type: library
- maturity: maintenance
- function: video-processing, machine-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, artificial-intelligence
- platform: python
- tags: video-representation, neural-fields, video-editing, video-to-video-translation, keypoint-tracking, research-code, cvpr-2024, pytorch, video, linux, gpu

## Member repositories
- ant-research/CoDeF (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:01.344488+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 2026-08-29T18:18:26.665630+00:00, confidence not recorded.
  - readme: https://github.com/ant-research/CoDeF (fetched 2026-08-28T04:09:01.344488+00:00, sha c398e6418558)
  - homepage: https://qiuyu96.github.io/CoDeF/ (fetched 2026-08-29T09:01:04.735098+00:00, sha 7f55a1f48f1a)
- Data as of 2026-08-30T08:39:29.467469+00:00.