# kazuto1011/deeplab-pytorch

PyTorch re-implementation of DeepLab v2 on COCO-Stuff / PASCAL VOC datasets

Repository: https://github.com/kazuto1011/deeplab-pytorch
Canonical: https://ross.abutalabs.com/products/deeplab-pytorch
Language: Python
License: MIT
License Family: permissive
Topics: pytorch, deeplab, semantic-segmentation, cocostuff, coco, voc
Last push: 2022-10-14T18:30:23+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3227, "days_push": 1419, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1100, forks 280 (observed 2026-08-28T04:03:35.314398+00:00)

## What it is
An unofficial PyTorch re-implementation of DeepLab v2 with a ResNet-101 backbone for semantic segmentation, supporting COCO-Stuff and PASCAL VOC datasets. It includes pretrained weights compatible with the official Caffe releases, optional DeepLab v3/v3+ variants, and torch.hub support.

## Use cases
- run semantic segmentation on images with deeplab in pytorch
- train deeplab v2 on pascal voc or coco-stuff
- download pretrained deeplab weights without building caffe
- reproduce deeplab v2 benchmark results like mean IoU
- use deeplab models via torch.hub
- get segmentation masks for coco-stuff 164k
- compare deeplab v2 v3 and v3+ with resnet-101 backbone

## When to choose
- you need a faithful PyTorch port of DeepLab v2 with verified benchmark numbers
- you want official pretrained weights without installing Caffe
- you work with COCO-Stuff or PASCAL VOC segmentation datasets
- you need a quick torch.hub-loadable segmentation model

## When to avoid
- you need the latest segmentation architectures like SegFormer or Mask2Former
- you want an actively developed project with frequent updates
- you need production-ready inference serving rather than research code
- you require the DeepLab v3/v3+ variants to be fully tested

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: deep-learning, computer-vision, image-processing, machine-learning
- platform: python, cross-platform
- tags: pytorch, deeplab, semantic-segmentation, resnet, coco-stuff, pascal-voc, torch-hub, pretrained-weights

## Member repositories
- kazuto1011/deeplab-pytorch (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:35.314398+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 2026-08-30T06:45:48.370182+00:00, confidence not recorded.
  - readme: https://github.com/kazuto1011/deeplab-pytorch (fetched 2026-08-28T04:03:35.314398+00:00, sha 621be5df1ad9)
- Data as of 2026-08-30T08:39:29.467469+00:00.