Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / yihui-he/KL-Loss issues and pull requests

#36 - negative loss

Issue - State: open - Opened by jeethesh-pai about 2 years ago

#35 - How to add KL Loss to YoloV5？

Issue - State: open - Opened by tuji-sjp almost 3 years ago

#34 - Question about code of KL loss

Issue - State: closed - Opened by Eaphan almost 3 years ago

#33 - How to output Box and Box std?

Issue - State: closed - Opened by Chao86 about 3 years ago - 1 comment

#32 - How to use this repo for detectron2

Issue - State: open - Opened by QuickLearner171998 about 3 years ago

#31 - Is KL loss also useful for one-stage detectors like retinanet?

Issue - State: closed - Opened by v-qjqs over 3 years ago - 2 comments

#30 - voc dataset

Issue - State: closed - Opened by lucs-C almost 4 years ago

#29 - about the meaning of Eq. 11

Issue - State: open - Opened by ShellingFord221 over 4 years ago

#28 - Lreg is always oscillatory and can‘t converge

Issue - State: open - Opened by DrewdropLife over 4 years ago

#27 - The value of the KL Loss

Issue - State: closed - Opened by pangsu0613 almost 5 years ago - 2 comments

#25 - hello:

Issue - State: closed - Opened by xiaoqiang-cheng almost 5 years ago

#24 - Questions about pre-trained weights

Issue - State: closed - Opened by JamesHao-ml almost 5 years ago - 1 comment

#23 - ap95 overwrite ap90 in cocoapi

Issue - State: open - Opened by ljquality about 5 years ago
Labels: bug

#22 - Sorry,How to add the KL-LOSS to faster-rcnn?I only find the fast-rcnn version

Issue - State: closed - Opened by LittleBoy7 about 5 years ago - 3 comments

#21 - implementation with KL-loss

Issue - State: closed - Opened by xiaowei-ui about 5 years ago - 3 comments

#20 - Any plans for PyTorch version with MMDetection?

Issue - State: open - Opened by LiJunnan1992 about 5 years ago - 7 comments
Labels: enhancement

#18 - To what part in the equations does each part of the loss correspond?

Issue - State: closed - Opened by JappaB about 5 years ago - 21 comments

#17 - AssertionError: Range subprocess failed (exit code: 1)

Issue - State: closed - Opened by MenGuangwen-CN-0411 about 5 years ago - 1 comment

#16 - p_i in Eq.11 doesn't included in the code?

Issue - State: closed - Opened by AppleBunnyy about 5 years ago - 1 comment

#15 - the loss_bbox is 0

Issue - State: closed - Opened by lg12170226 about 5 years ago - 2 comments

#14 - KL

Issue - State: closed - Opened by github-luffy about 5 years ago - 3 comments

#13 - KL-Loss very large

Issue - State: closed - Opened by EternityZY about 5 years ago - 5 comments

#12 - train err

Issue - State: closed - Opened by lg12170226 over 5 years ago

#11 - What do the bbox_pred and bbox_targets mean?

Issue - State: closed - Opened by JoyHuYY1412 over 5 years ago - 1 comment

#10 - could you also show the position of softNMS and variance-voting?

Issue - State: closed - Opened by JoyHuYY1412 over 5 years ago - 3 comments

#8 - What does l12mask work for?

Issue - State: closed - Opened by Ocelot7777 over 5 years ago

#7 - what do the 'bbox_inside_weights' and 'bbox_outside_weights' mean?

Issue - State: closed - Opened by JoyHuYY1412 over 5 years ago - 2 comments

#6 - 请问一下，这个里面方差具体是怎么预测的？

Issue - State: closed - Opened by shenghsiaowong over 5 years ago - 1 comment

#5 - Code location of KL-Loss

Issue - State: closed - Opened by DCNSW over 5 years ago - 2 comments

#4 - Extra term in the KL-loss code and the NaN problem

Issue - State: closed - Opened by huiqu18 over 5 years ago - 2 comments

#3 - KL-loss seems could be negative?

Issue - State: closed - Opened by memoiry over 5 years ago - 1 comment

#2 - What's the scale of two loss when training KL loss

Issue - State: closed - Opened by hding2455 over 5 years ago - 1 comment

#1 - Fix indent

Pull Request - State: closed - Opened by zhangliliang over 5 years ago