-
Notifications
You must be signed in to change notification settings - Fork 9.4k
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
Add tta to HTC and Cascade RCNN #1251
Conversation
Nice work! Could you fix the linting errors? |
I will try it. This is the first time I've contributed code to the open source community, and I've ignored the importance of the code format. |
Have you tried multi-scale training and how much mAP improvement you get ? Thanks! @donglee-afar |
I haven't try it on coco dataset, but in my experience on Kaggle and some other competition dataset, it usually can boost about 0.5~1.5% mAP |
This looks promising... Will this work on retinanet? Is it only active as is in the "test" mode or also used with the API for single images? |
Thanks a lot, I tried the multi-scale training but I find the mAP get worse. So do you think multi-scale training is necessary when I already use FPN ? @donglee-afar 另外能否留个企鹅啥的 |
qq:1955662900 |
scale_factor, flip) | ||
mask_rois = bbox2roi([_bboxes]) | ||
mask_roi_extractor = self.mask_roi_extractor[-1] | ||
mask_feats = mask_roi_extractor( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for your work. This two statements seems redundant.
mask_rois = bbox2roi([_bboxes]) | ||
mask_feats = self.mask_roi_extractor[-1]( | ||
x[:len(self.mask_roi_extractor[-1].featmap_strides)], | ||
mask_rois) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
mask_feats
seems redundant.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You are right, mask_feats does not need to be calculated here in cascadercnn because there is no semantic_feat to be add.
I have delete it and test it again, results is the same as what i posted before.
* add tta to HTC and Caccade RCNN * format file with yapf * fix import error with isort * Update htc.py * Update cascade_rcnn.py * fix bug * delete some redundant codes
* Vision: Error if invalid kwargs specified * Fixed vision unit tests * Fixed vision docs * Fix vision HPO
Add tta to HTC and Cascade RCNN
write with the similar tta scheme of fasterrcnn in this repo
Here are the experiment results
Cascade Mask R-CNN
Hybrid Task Cascade (HTC)
experiment with the config file: