# msracver/Relation-Networks-for-Object-Detection

Relation Networks for Object Detection

Repository: https://github.com/msracver/Relation-Networks-for-Object-Detection
Canonical: https://ross.abutalabs.com/products/relation-networks-for-object-detection
Language: Python
License: MIT
License Family: permissive
Last push: 2021-09-27T04:22:40+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": 3002, "days_push": 1801, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1104, forks 187 (observed 2026-08-28T04:03:36.261755+00:00)

## What it is
Official MXNet implementation of the CVPR 2018 paper 'Relation Networks for Object Detection', which adds an attention-based relation module to Faster R-CNN to replace NMS and improve detection accuracy. It includes training and evaluation code on the COCO dataset with ResNet-101 backbones.

## Use cases
- reproduce relation network object detection results on COCO
- train Faster R-CNN with a learned relation module instead of NMS
- study attention-based duplicate removal for object detection
- compare learn NMS versus soft-NMS baselines
- build object detection research on top of a published CVPR 2018 model

## When to choose
- you need the official reference implementation of the Relation Networks paper
- you work with MXNet and want a research-grade object detection codebase
- you want to experiment with learned NMS and relation modules in Faster R-CNN

## When to avoid
- you need a production object detection system with active maintenance
- you prefer PyTorch or TensorFlow tooling
- you need support for recent CUDA versions or modern GPU hardware

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: object-detection, mxnet, relation-networks, faster-rcnn, nms, research-code, cvpr-2018, linux, gpu

## Member repositories
- msracver/Relation-Networks-for-Object-Detection (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:36.261755+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 2026-08-30T06:45:28.982495+00:00, confidence not recorded.
  - readme: https://github.com/msracver/Relation-Networks-for-Object-Detection (fetched 2026-08-28T04:03:36.261755+00:00, sha 47dd5a979bd8)
- Data as of 2026-08-30T08:39:29.467469+00:00.