# NVIDIA/vid2vid

Pytorch implementation of our method for high-resolution (e.g. 2048x1024) photorealistic video-to-video translation.

Repository: https://github.com/NVIDIA/vid2vid
Canonical: https://ross.abutalabs.com/products/vid2vid
Language: Python
License: NOASSERTION
License Family: other
Last push: 2022-05-17T05:09:30+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 100
- inputs: {"age_days": 2941, "days_push": 1569, "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 8692, forks 1207 (observed 2026-08-28T04:10:24.258084+00:00)

## What it is
A PyTorch implementation of NVIDIA's video-to-video synthesis method for generating high-resolution (e.g., 2048x1024) photorealistic videos from semantic label maps, edge maps, or pose sequences. It is the official research code accompanying the NeurIPS 2018 paper 'Video-to-Video Synthesis'.

## Use cases
- turning semantic segmentation maps into photorealistic street view videos
- synthesizing talking faces from facial edge maps
- generating human body motion videos from pose keypoints
- frame prediction for video synthesis
- reproducing the vid2vid NeurIPS 2018 paper results
- training video-to-video translation models on custom datasets

## When to choose
- you need to translate semantic labels, edges, or poses into photorealistic video
- you want the reference implementation of the vid2vid paper for research
- you have an NVIDIA GPU and can work with PyTorch 0.4-era code

## When to avoid
- you need a maintained production-ready video generation tool
- you want modern PyTorch versions or recent GPU support out of the box
- you need general-purpose video editing rather than conditional video synthesis

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, video-processing, image-processing
- domain: deep-learning, computer-vision, machine-learning
- platform: python
- tags: video-to-video-translation, gan, pytorch, image-to-image-translation, research-code, neurips-2018, video, linux, macos, gpu

## Member repositories
- NVIDIA/vid2vid (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:24.258084+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 2026-08-29T17:25:34.587384+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA/vid2vid (fetched 2026-08-28T04:10:24.258084+00:00, sha b61f34b02ee9)
- Data as of 2026-08-30T08:39:29.467469+00:00.