# KaihuaTang/Scene-Graph-Benchmark.pytorch

A new codebase for popular Scene Graph Generation methods (2020). Visualization & Scene Graph Extraction on custom images/datasets are provided. It's also a PyTorch implementation of paper “Unbiased Scene Graph Generation from Biased Training CVPR 2020”

Repository: https://github.com/KaihuaTang/Scene-Graph-Benchmark.pytorch
Canonical: https://ross.abutalabs.com/products/scene-graph-benchmarkpytorch
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2026-01-06T02:25:00+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 60, release rhythm 35, longevity 100
- inputs: {"age_days": 2386, "days_push": 240, "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 1195, forks 236 (observed 2026-08-28T04:03:56.848206+00:00)

## What it is
A PyTorch codebase for Scene Graph Generation (SGG) built on maskrcnn-benchmark, implementing methods from the CVPR 2020 paper 'Unbiased Scene Graph Generation from Biased Training'. It supports training, benchmarking with standard SGG metrics, and extracting/visualizing scene graphs on custom images.

## Use cases
- generate scene graphs from images
- benchmark scene graph generation models
- run unbiased SGG from biased training
- visualize detected scene graphs on custom images
- train relationship prediction models in PyTorch
- evaluate SGG with Recall@K and mean Recall metrics
- extract scene graphs from a custom dataset

## When to choose
- you need a standard benchmark codebase for scene graph generation research
- you want to reproduce or extend CVPR 2020 unbiased SGG methods
- you need scene graph extraction and visualization on your own images

## When to avoid
- you need production-ready object detection rather than research SGG code
- you require a modern maintained framework rather than a 2020-era PyTorch codebase
- you work outside Python/PyTorch environments

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, computer-vision, image-processing, benchmarking, data-visualization
- domain: computer-vision, deep-learning, machine-learning, image-processing
- platform: python
- tags: scene-graph-generation, pytorch, cvpr-2020, relationship-prediction, maskrcnn-benchmark, sgg, research-code, linux, gpu

## Member repositories
- KaihuaTang/Scene-Graph-Benchmark.pytorch (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:56.848206+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 2026-08-30T06:22:05.066947+00:00, confidence not recorded.
  - readme: https://github.com/KaihuaTang/Scene-Graph-Benchmark.pytorch (fetched 2026-08-28T04:03:56.848206+00:00, sha 36ab17ce4af7)
- Data as of 2026-08-30T08:39:29.467469+00:00.