WXinlong/SOLO

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1758 stars · 306 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Official PyTorch implementation of SOLO and SOLOv2, box-free fully convolutional methods for instance segmentation published at ECCV 2020 and NeurIPS 2020. Built on mmdetection and includes pretrained COCO models.

Use cases

• segment individual object instances in images
• run box-free instance segmentation with pretrained COCO models
• reproduce SOLO/SOLOv2 research results
• fine-tune an instance segmentation model on custom data
• compare instance segmentation model performance on COCO

When to choose

• you need direct, box-free instance mask prediction
• you want high-quality object boundary masks
• you are building on mmdetection and want a state-of-the-art segmentation model
• you need a fast lightweight segmentation model (SOLOv2 at ~31 FPS on V100)

When to avoid

• you need panoptic or semantic segmentation instead of instance segmentation
• you want a production-maintained framework rather than research code
• you need a license permitting commercial use (license is non-standard)
• you prefer a non-PyTorch stack without mmdetection dependencies

Facets

library · maturity maintenance

computer-vision image-processing machine-learning deep-learning computer-vision image-processing machine-learning deep-learning python instance-segmentation object-detection pytorch mmdetection coco research-code solo solov2 box-free fcn gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="WXinlong/SOLO")