# nbei/Deep-Flow-Guided-Video-Inpainting

pytorch implementation for "Deep Flow-Guided Video Inpainting"(CVPR'19)

Repository: https://github.com/nbei/Deep-Flow-Guided-Video-Inpainting
Canonical: https://ross.abutalabs.com/products/deep-flow-guided-video-inpainting
Homepage: https://nbei.github.io/video-inpainting.html
Language: Python
License: MIT
License Family: permissive
Last push: 2022-12-08T05:52:59+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2717, "days_push": 1364, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2375, forks 438 (observed 2026-08-28T04:06:42.017163+00:00)

## What it is
A PyTorch implementation of the CVPR 2019 paper 'Deep Flow-Guided Video Inpainting', which fills missing regions in videos by completing optical flow fields and propagating pixels guided by them. It bundles a video inpainting tool (DFVI), a modified FlowNet2 for flow extraction, and a reimplemented DeepFillv1 image inpainting network.

## Use cases
- remove objects from videos
- fill in missing regions in video frames
- video inpainting with masks
- complete occluded areas in footage
- research baseline for flow-guided video inpainting
- inpaint fixed regions in DAVIS-style videos

## When to choose
- you need to remove objects or fill holes in video sequences with temporal coherence
- you want a research-grade reference implementation of flow-guided video inpainting
- you have a GPU and can work with PyTorch 0.4-era code

## When to avoid
- you need a maintained, modern PyTorch pipeline or easy pip installation
- you want real-time or production video editing without GPU setup
- you only need single-image inpainting without video context

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, video-processing, image-processing, computer-vision
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: video-inpainting, optical-flow, pytorch, object-removal, cvpr-2019, flownet2, research-code, video, linux, docker, gpu

## Member repositories
- nbei/Deep-Flow-Guided-Video-Inpainting (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:42.017163+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 2026-08-30T02:35:03.494638+00:00, confidence not recorded.
  - readme: https://github.com/nbei/Deep-Flow-Guided-Video-Inpainting (fetched 2026-08-28T04:06:42.017163+00:00, sha a291d17b8758)
  - homepage: https://nbei.github.io/video-inpainting.html (fetched 2026-08-29T10:16:04.840446+00:00, sha 072651bc878e)
- Data as of 2026-08-30T08:39:29.467469+00:00.