mmdet 3.3.0


OpenMMLab Detection Toolbox and Benchmark

OpenMMLab Detection Toolbox and Benchmark

Stars: 27593, Watchers: 27593, Forks: 9178, Open Issues: 1512

The open-mmlab/mmdetection repo was created 5 years ago and the last code push was 13 hours ago.
The project is extremely popular with a mindblowing 27593 github stars!

How to Install mmdet

You can install mmdet using pip

pip install mmdet

or add it to a project with poetry

poetry add mmdet

Package Details

MMDetection Contributors
Apache License 2.0
GitHub Repo:


No  mmdet  pypi packages just yet.


A list of common mmdet errors.

Code Examples

Here are some mmdet code examples and snippets.

GitHub Issues

The mmdet package has 1512 open issues on GitHub

  • How to export model to onnx without post-processing
  • The code may have some bugs
  • [CodeCamp2023-489]Add new configuration files for htc in mmdetection
  • AttributeError: 'int' object has no attribute 'type'
  • [CodeCamp2023-500]add large_image_demo
  • There is a misspelling in test_time_augs/
  • [CodeCamp2023-476]Add new configuration files for QDTrack algorithm in mmdetection
  • [bug] one of the variables needed for gradient computation has been modified by an inplace operation
  • [MMdetection 2] How to apply data augmentation in test_data ?
  • Support OneFormer
  • Update
  • About the iou of rtmdet
  • Update Instance segmentation Tutorial
  • How to use MMdetection3.0 to train one's own grayscale image data?
  • Features/foggy cityscapes oracle

See more issues on GitHub

Related Packages & Articles

ultralytics 8.1.45

Ultralytics YOLOv8 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.

nncf 2.9.0

Neural Networks Compression Framework

stardist 0.8.5

StarDist is a Python package designed for object detection with star-convex shapes, particularly useful in microscopy. It provides implementations for both 2D and 3D images and uses a model trained to predict distances to the object boundary along fixed rays and object probabilities. These predictions produce an overcomplete set of candidate polygons for a given image, with the final result obtained via non-maximum suppression.

StarDist is compatible with Python 3.6 to 3.10, requires TensorFlow, and provides pre-trained models for 2D images and example workflows via Jupyter notebooks, making it a versatile tool for cell detection and segmentation in microscopy.

detectors 0.1.11

This is a comprehensive library for generalized Out-of-Distribution (OOD) detection research. It provides over 20 detection methods, evaluation pipelines, OOD datasets, and model architectures integrated with timm. Additionally, it offers fast OOD evaluation metrics, multi-layer detection methods, and pipelines for open set recognition and covariate drift detection.