IS-Net is a deep learning model designed to provide interactive image segmentation capabilities. The model allows users to refine segmentation masks through user interactions, making it highly effective for tasks that require precise and detailed segmentation results.

Usage

Perform mask generation with BritishWerewolf/IS-Net.

Example

javascript

import { AutoModel, AutoProcessor, RawImage } from '@huggingface/transformers';

const img_url = 'https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png';
const image = await RawImage.read(img_url);

const processor = await AutoProcessor.from_pretrained('BritishWerewolf/IS-Net');
const processed = await processor(image);

const model = await AutoModel.from_pretrained('BritishWerewolf/IS-Net', {
    dtype: 'fp32',
});

const output = await model({ input: processed.pixel_values });
// {
//   mask: Tensor {
//     dims: [ 1, 1024, 1024 ],
//     type: 'uint8',
//     data: Uint8Array(1048576) [ ... ],
//     size: 1048576
//   }
// }

Inference

To use the model for inference, you can follow the example provided above. The AutoProcessor and AutoModel classes from the transformers library make it easy to load the model and processor.

Credits

rembg for the ONNX model.
The authors of the original IS-Net model can be credited at https://github.com/xuebinqin/DIS.

Licence

This model is licensed under the Apache License 2.0 to match the original IS-Net model.