Ir para o conteúdo
ModelListadosExterno
Browse files

Isnet General Use Onnx

ONNX builds of ISNet (general use) from xuebinqin/DIS (Apache-2.0), packaged for Transformers.js.

Fontehuggingface.co/Ko033/isnet-general-use-onnx/tree/5349b617911fd60c619b52f32e2b593517b78df3
TipoOtherVersãoRevision 5349b61Bnota

Fonte do modelo

Tipo
Other
Tamanho do arquivo
43,8–84,1 MB
Versão
vRevision 5349b61
Atualizado
9 de out. de 2026
Downloads
763
Fonte
Hugging Face

Descrição da fonte

isnet-general-use (ONNX, fp16 / q8)

ONNX builds of ISNet (general use) from xuebinqin/DIS (Apache-2.0), packaged for Transformers.js.

  • Source ONNX export: the isnet-general-use.onnx distributed by the rembg project (MIT); original weights and architecture by Xuebin Qin et al. (DIS, Apache-2.0).
  • Modifications: the 11 auxiliary graph outputs (side stages d2-d6 and intermediate feature taps) were removed, keeping only output_image (stage-1 sigmoid probability map). fp16 and dynamic-int8 (q8) variants were generated with the official Transformers.js conversion scripts.

I/O

  • Input input_image: float32 [1, 3, 1024, 1024], RGB, normalized as x / 255 - 0.5 (mean 0.5, std 1.0).
  • Output output_image: float32 [1, 1, 1024, 1024], sigmoid probabilities. Min-max normalize per image for an alpha matte (as done by rembg).

Files

  • onnx/model.onnx - fp32 (aux outputs stripped)
  • onnx/model_fp16.onnx - fp16
  • onnx/model_quantized.onnx - dynamic int8 (q8)

Fontes

1 fonte
Hugging FacePrimárioVersão 5349b617911fd60c619b52f32e2b593517b78df3
huggingface.co/Ko033/isnet-general-use-onnx/tree/5349b617911fd60c619b52f32e2b593517b78df3

Verificado 9 de out.

Download files

Artefatos de modelo

2 artefatos

onnx/model_fp16.onnx

onnx · 84,1 MB · SHA-256 0857167263ad…d32c · Hugging Face

Download file

onnx/model_quantized.onnx

onnx · 43,8 MB · SHA-256 5039225b9a4a…03a6 · Hugging Face

Download file

Trechos de fonte

3 trechos
Readme31 de jul.huggingface.co/Ko033/isnet-general-use-onnx

--- license: apache-2.0 library_name: transformers.js pipeline_tag: image-segmentation tags: - onnx - background-removal - dichotomous-image-segmentation --- # isnet-general-use (ONNX, fp16 / q8) ONNX builds of **ISNet (general use)** from [xuebinqin/DIS](https://github.com/xuebinqin/DIS) (Apache-2.0), packaged for [Transformers.js](https://huggingface.co/docs/transformers.js). - Source ONNX export: the `isnet-general-use.onnx` distributed by the [rembg](https://github.com/danielgatis/rembg) project (MIT); original weights and architecture by Xuebin Qin et al. (DIS, Apache-2.0). - Modifications: the 11 auxiliary graph outputs (side stages d2-d6 and intermediate feature taps) were removed, keeping only `output_image` (stage-1 sigmoid probability map). fp16 and dynamic-int8 (q8) variants were generated with the official Transformers.js conversion scripts. ## I/O - Input `input_image`: float32 `[1, 3, 1024, 1024]`, RGB, normalized as `x / 255 - 0.5` (mean 0.5, std 1.0). - Output `output_image`: float32 `[1, 1, 1024, 1024]`, sigmoid probabilities. Min-max normalize per image for an alpha matte (as done by rembg). ## Files - `onnx/model.onnx` - fp32 (aux outputs stripped) - `onnx/model_fp16.onnx` - fp16 - `onnx/model_quantized.onnx` - dynamic int8 (q8)

Note9 de out.huggingface.co/Ko033/isnet-general-use-onnx

Source context: 763 downloads · 0 likes · Pipeline image-segmentation · Library transformers.js · Repo Ko033/isnet-general-use-onnx

Note31 de jul.huggingface.co/Ko033/isnet-general-use-onnx

Source context: 76 downloads · 0 likes · Pipeline image-segmentation · Library transformers.js · Repo Ko033/isnet-general-use-onnx