open-mmlab/mmcv

Health v2 · maintenance only

Adoption not part of the score

6470 stars · 1771 forks observed · 2026-08-28

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

MMCV is the foundational computer vision library for the OpenMMLab ecosystem, providing image/video I/O, data transformations, and CUDA operators built on PyTorch. It serves as the base dependency for OpenMMLab projects like MMDetection and MMSegmentation.

Use cases

• build computer vision models on pytorch
• image and video io and preprocessing
• run cuda ops for detection models
• data augmentation transforms for vision training
• foundation library for mmdetection and mmsegmentation

When to choose

• you use OpenMMLab toolboxes like MMDetection, MMSegmentation, or MMPose
• you need optimized CUDA vision operators with PyTorch
• you want a shared set of image/video transforms and I/O utilities

When to avoid

• you need a full training engine - use MMEngine instead
• you want a standalone high-level vision framework rather than a foundation library
• you don't use PyTorch

Facets

library · maturity stable

image-processing machine-learning deep-learning data-visualization computer-vision deep-learning machine-learning windows python pytorch opencv computer-vision-foundation openmmlab data-transformations cuda-ops linux macos gpu cuda

Member repositories

For agents

markdown · JSON · MCP: product_card(name="open-mmlab/mmcv")