# jiupinjia/SkyAR

Official Pytorch implementation of the preprint paper "Castle in the Sky: Dynamic Sky Replacement and Harmonization in Videos", in arXiv:2010.11800.

Repository: https://github.com/jiupinjia/SkyAR
Canonical: https://ross.abutalabs.com/products/skyar
Homepage: https://jiupinjia.github.io/skyar/
Language: Python
License Family: other
Topics: dynamic-sky-replacement, sky-matting, harmonization, skyar
Last push: 2022-08-04T12:29:18+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": 2141, "days_push": 1490, "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 2025, forks 245 (observed 2026-08-28T04:06:06.737605+00:00)

## What it is
SkyAR is the official PyTorch implementation of the paper 'Castle in the Sky: Dynamic Sky Replacement and Harmonization in Videos'. It performs vision-based, real-time sky replacement and harmonization in videos using sky matting, motion estimation, and image blending.

## Use cases
- replace the sky in a video with a dramatic background
- add a floating castle or galaxy sky to my footage
- change weather and lighting in outdoor videos
- augment driving dataset videos with different sky conditions
- run real-time sky replacement on handheld or dashcam video

## When to choose
- you need automatic, interaction-free sky replacement in videos
- you want to reproduce the arXiv 2010.11800 paper results
- you need weather/lighting translation for data augmentation
- you have a GPU and want real-time processing

## When to avoid
- you need a permissive or commercial license (it is CC BY-NC-SA)
- you only need static photo sky replacement
- you need a production-supported tool with active maintenance
- you cannot download external pretrained checkpoints

## Facets
- artifact type: library
- maturity: maintenance
- function: video-processing, image-processing, computer-vision, machine-learning, deep-learning
- domain: computer-vision, image-processing, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: sky-replacement, sky-matting, video-editing, pytorch, research-code, non-commercial-license, video, gpu, linux

## Member repositories
- jiupinjia/SkyAR (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:06.737605+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 2026-08-30T02:59:45.613331+00:00, confidence not recorded.
  - readme: https://github.com/jiupinjia/SkyAR (fetched 2026-08-28T04:06:06.737605+00:00, sha f9bde5d41c5a)
  - homepage: https://jiupinjia.github.io/skyar/ (fetched 2026-08-29T10:39:56.212573+00:00, sha 34efe03600a0)
- Data as of 2026-08-30T08:39:29.467469+00:00.