dmlc/gluon-cv

Health v2 · maintenance only

Adoption not part of the score

5916 stars · 1205 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

GluonCV is a deep learning toolkit providing state-of-the-art computer vision model implementations with 170+ pre-trained models. It supports both MXNet and PyTorch backends and covers image classification, object detection, segmentation, pose estimation, and action recognition.

Use cases

• run image classification with pretrained models
• train an object detection model on custom data
• semantic segmentation of images
• human pose estimation
• video action recognition
• reproduce SOTA results from research papers
• export models for mobile deployment

When to choose

• you need pretrained computer vision models with a simple one-line API
• you want training scripts to reproduce published SOTA results
• you work with MXNet or want dual PyTorch/MXNet support

When to avoid

• you need actively maintained tooling - MXNet is retired and development has shifted to AutoGluon and other projects
• you need tasks beyond the supported vision applications
• you want the latest model architectures, which appear faster in timm or MMDetection

Facets

library · maturity maintenance

machine-learning deep-learning image-processing computer-vision computer-vision deep-learning machine-learning image-processing python windows model-zoo pretrained-models mxnet pytorch object-detection semantic-segmentation pose-estimation action-recognition image-classification linux macos gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="dmlc/gluon-cv")