# yangxue0827/RotationDetection

This is a tensorflow-based rotation detection benchmark, also called AlphaRotate.

Repository: https://github.com/yangxue0827/RotationDetection
Canonical: https://ross.abutalabs.com/products/rotationdetection
Homepage: https://rotationdetection.readthedocs.io/
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2024-11-07T16:09:15+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2137, "days_push": 664, "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 1118, forks 179 (observed 2026-08-28T04:03:39.160784+00:00)

## What it is
AlphaRotate is a TensorFlow-based benchmark and toolbox for rotated (oriented) object detection, implementing detectors such as R2CNN, RetinaNet-based rotation variants, and related techniques with support for standard backbones and FPN-style necks. It ships with training and evaluation pipelines for widely used rotated detection datasets including DOTA, HRSC2016, ICDAR2015/2017, MSRA-TD500, UCAS-AOD, and Total-Text, and is maintained as an academic research benchmark by Shanghai Jiao Tong University.

## Use cases
- detect objects with rotated bounding boxes in aerial or remote sensing images
- train a rotated object detector on the DOTA dataset
- benchmark oriented object detection algorithms like R2CNN
- perform scene text detection with arbitrary orientations on ICDAR datasets
- compare rotated detection models under a single TensorFlow framework
- reproduce research results for oriented bounding box detection

## When to choose
- you work with TensorFlow and need rotated/oriented object detection
- your research targets remote sensing, aerial imagery, or rotated scene text
- you want a unified benchmark to compare multiple rotation detection algorithms
- you need pretrained or reproducible baselines on DOTA, HRSC2016, or ICDAR benchmarks

## When to avoid
- your stack is PyTorch-first (consider MMRotate or MMDetection-based alternatives)
- you only need standard horizontal bounding box detection
- you need a production-ready, commercially supported detection service rather than a research benchmark
- you require lightweight inference on edge devices with minimal dependencies

## Facets
- artifact type: framework
- maturity: active
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning, artificial-intelligence, image-processing
- platform: python, cross-platform
- tags: object-detection, rotated-bounding-box, oriented-object-detection, remote-sensing, aerial-imagery, text-detection, benchmark, dota-dataset, tensorflow, alpharotate, research, gpu, linux

## Member repositories
- yangxue0827/RotationDetection (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:39.160784+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 2026-08-30T06:41:50.807149+00:00, confidence not recorded.
  - readme: https://github.com/yangxue0827/RotationDetection (fetched 2026-08-28T04:03:39.160784+00:00, sha 2bb520203acf)
- Data as of 2026-08-30T08:39:29.467469+00:00.