Skip to content

A simple implementation of Faster-RCNN with minibatch, multi-gpu. FPN will be added soon

Notifications You must be signed in to change notification settings

yfji/pytorch-faster-rcnn

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Faster-RCNN using PyTorch

This code contains the training and test of Faster-RCNN

Contents

Train

  • Training process is implemented in train_frcnn.py
  • You need to realize your own data_loader in roidb. deepdrive_data_loader.py is implemented for DeepDrive of AIChallenge dataset. You can also change it into VOC data loader or COCO data loader by overwriting get_minibatch()
  • More details in train_frcnn.py. The default number of epochs is 50, and the learning rate decays by 0.1 at epoch 30. You can change it in train()
  • The first 100 images containing the foreground anchors are visualized iin 'vis_anchors'. This is a test for generating anchors.

Test

  • Test is implemented in inference.py
  • Put the testing image in to folder 'images' and modify inference.py
  • You can visualize proposals output by RPN by uncommenting draw_proposals at line 181

Results

  • Detection Load failed
  • Proposals Load failed

About

A simple implementation of Faster-RCNN with minibatch, multi-gpu. FPN will be added soon

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published