# timojl/clipseg

This repository contains the code of the CVPR 2022 paper "Image Segmentation Using Text and Image Prompts".

Repository: https://github.com/timojl/clipseg
Canonical: https://ross.abutalabs.com/products/clipseg
Language: Python
License: NOASSERTION
License Family: other
Last push: 2024-01-05T10:36:22+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": 1745, "days_push": 971, "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 1338, forks 121 (observed 2026-08-28T04:04:25.812732+00:00)

## What it is
CLIPSeg is a Python implementation of the CVPR 2022 paper 'Image Segmentation Using Text and Image Prompts', enabling zero-shot segmentation of images based on arbitrary text queries or prompt images with masks. It provides pre-trained model weights and has been integrated into the HuggingFace Transformers library.

## Use cases
- segment objects in an image using a text prompt
- zero-shot image segmentation without training
- create segmentation masks from a reference image with a highlighted region
- run text-guided segmentation in a notebook
- fine-grained segmentation predictions with pre-trained weights

## When to choose
- you need to segment arbitrary objects described by text without training a model
- you want lightweight pre-trained weights for zero-shot segmentation
- you prefer a model already supported in HuggingFace Transformers

## When to avoid
- you need fully supervised, dataset-specific segmentation with maximum accuracy
- you require a permissive license for the model weights (weights have a custom license)
- you need real-time segmentation on CPU-only hardware

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, computer-vision, image-processing, deep-learning
- domain: computer-vision, image-processing, machine-learning, deep-learning
- platform: python
- tags: image-segmentation, zero-shot, clip, text-prompted-segmentation, cvpr-2022, huggingface-transformers, gpu

## Member repositories
- timojl/clipseg (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:25.812732+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 2026-08-30T04:44:01.475450+00:00, confidence not recorded.
  - readme: https://github.com/timojl/clipseg (fetched 2026-08-28T04:04:25.812732+00:00, sha 8915b21a666d)
- Data as of 2026-08-30T08:39:29.467469+00:00.