Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fast rcnn #5

Open
wants to merge 24 commits into
base: master
Choose a base branch
from
Open

Fast rcnn #5

wants to merge 24 commits into from

Conversation

fmassa
Copy link
Owner

@fmassa fmassa commented Aug 23, 2015

Implements fast-rcnn detection framework. This PR includes

  • a simple and efficient lua implementation for ROIPooling based on SpatialAdaptiveMaxPooling
  • train and test code
  • an image-centric BatchProvider

Some code refactor is needed before this PR gets merged.
For the moment, there is no bounding-box regression.

@fmassa
Copy link
Owner Author

fmassa commented Sep 2, 2015

Just a heads up on the status of this PR.
Using Caffenet, without bbox regression, this code gives a mAP of around 50.3 on Pascal VOC 2007.
Once I figure out what's holding back the ~1/2 points from original Fast-RCNN, I'll clean it and merge it.

szagoruyko and others added 4 commits September 22, 2015 12:04
so the  BatchProviderROI inherited fields can be properly saved by torch.save
Fix BatchProviderROI incorrect super constructor invokation
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants