MMDetection
v2.6.0

Get Started

  • Prerequisites
  • Installation
  • Verification
  • Benchmark and Model Zoo

Quick Run

  • 1: Inference and train with existing models and standard datasets
  • 2: Train with customized datasets

Tutorials

  • Tutorial 1: Learn about Configs
  • Tutorial 2: Customize Datasets
  • Tutorial 3: Customize Data Pipelines
  • Tutorial 4: Customize Models
  • Tutorial 5: Customize Runtime Settings
  • Tutorial 6: Customize Losses
  • Tutorial 7: Finetuning Models

Useful Tools and Scripts

  • Log Analysis
  • Visualization
  • Error Analysis
  • Model Complexity
  • Model conversion
  • Dataset Conversion
  • Miscellaneous

Notes

  • Compatibility with MMDetection 1.x
  • Projects based on MMDetection
  • Changelog
  • Trouble Shooting

API Reference

  • API Reference
MMDetection
  • Docs »
  • <no title>
  • Edit on GitHub

  • Tutorial 1: Learn about Configs
    • Config File Structure
    • Config Name Style
    • An Example of Mask R-CNN
    • FAQ
  • Tutorial 2: Customize Datasets
    • Support new data format
    • Customize datasets by dataset wrappers
    • Modify Dataset Classes
  • Tutorial 3: Customize Data Pipelines
    • Design of Data pipelines
    • Extend and use custom pipelines
  • Tutorial 4: Customize Models
    • Develop new components
  • Tutorial 5: Customize Runtime Settings
    • Customize optimization settings
    • Customize training schedules
    • Customize workflow
    • Customize hooks
  • Tutorial 6: Customize Losses
    • Computation pipeline of a loss
    • Tweaking loss
    • Weighting loss (step 2)
  • Tutorial 7: Finetuning Models
    • Inherit base configs
    • Modify head
    • Modify dataset
    • Modify training schedule
    • Use pre-trained model
Next Previous

© Copyright 2018-2020, OpenMMLab Revision bd3306f5.

Built with Sphinx using a theme provided by Read the Docs.