# whai362/PVT

Official implementation of PVT series

Repository: https://github.com/whai362/PVT
Canonical: https://ross.abutalabs.com/products/pvt
Language: Python
License: Apache-2.0
License Family: permissive
Topics: transformer, backbone, pvt, detection, segmentation, pvtv2
Last push: 2022-10-27T08:47:14+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": 2017, "days_push": 1406, "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 1900, forks 252 (observed 2026-08-28T04:05:51.525812+00:00)

## What it is
Official PyTorch implementation of the Pyramid Vision Transformer (PVTv1 and PVTv2) model series, providing pretrained backbone networks for computer vision. It includes model weights and configs for image classification, object detection, and semantic segmentation tasks.

## Use cases
- use a vision transformer backbone for image classification on ImageNet
- pretrained PVT backbone for object detection with Mask R-CNN or RetinaNet
- semantic segmentation models using PVT as encoder
- compare PVTv2 against Swin Transformer backbones
- extract multi-scale features from images for downstream vision tasks
- fine-tune a pyramid vision transformer on custom datasets

## When to choose
- you need a well-tested transformer backbone for detection or segmentation with mmdetection/mmsegsupport
- you want pretrained ImageNet weights for a pyramid-style vision transformer
- you are reproducing or building on the PVT papers

## When to avoid
- you need a general-purpose training framework rather than specific backbone models
- you require a CNN backbone like ResNet instead of transformers
- you need actively maintained code with recent updates, as development has slowed

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: vision-transformer, backbone, pyramid-vision-transformer, object-detection, semantic-segmentation, image-classification, pretrained-models, pytorch

## Member repositories
- whai362/PVT (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:51.525812+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 2026-08-30T03:12:05.078955+00:00, confidence not recorded.
  - readme: https://github.com/whai362/PVT (fetched 2026-08-28T04:05:51.525812+00:00, sha d63354912a6d)
- Data as of 2026-08-30T08:39:29.467469+00:00.