https://github.com/aoirint/anime-segmentation-docker

SkyTNT/anime-segmentation in Docker
https://github.com/aoirint/anime-segmentation-docker

Last synced: about 1 month ago
JSON representation

SkyTNT/anime-segmentation in Docker

• Host: GitHub
• URL: https://github.com/aoirint/anime-segmentation-docker
• Owner: aoirint
• License: mit
• Created: 2023-02-25T08:09:53.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-07-10T04:18:36.000Z (6 months ago)
• Last Synced: 2024-07-10T06:41:48.125Z (6 months ago)
• Language: Dockerfile
• Homepage: https://hub.docker.com/r/aoirint/anime_segmentation
• Size: 42 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Anime Segmentation in Docker

- 

## Environments

- Ubuntu 20.04 or later

- [Docker Engine](https://docs.docker.com/engine/install/ubuntu/) 23.0 or later

- [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html)

## Usage

### 1. Build Docker image

```shell

docker build . -t aoirint/anime_segmentation

```

### 2. Download Model

- 

```shell

#  Create a model directory (UID:GID = 1000:1000)

mkdir model

wget 'https://huggingface.co/skytnt/anime-seg/resolve/a0a563c41338cbe0d23dfb4bfc3e243c518e5768/isnetis.ckpt'

echo '2c8f6b9a77386c54dcdbf55b6c917108c4bdf4328abca9152c7bce5727b74d18  isnetis.ckpt' | sha256sum -c -

```

### 3. Run Inference

```shell

# Create a input / output directory (UID:GID = 1000:1000)

mkdir data

mkdir out

docker run --rm --gpus all -v "./model:/model" -v "./data:/data" -v "./out:/out" aoirint/anime_segmentation --net isnet_is --ckpt "/model/isnetis.ckpt" --data "/data" --out "/out" --img-size 1024 --only-matted

```