ColdWind
Me also waiting for a life saver, and before that, using `locality_aware_nms.py` instead of `lanms` might be a sloowwwww solution
in `lanms/`, edit the `Makefile`, and the script looks like: ```shell CXXFLAGS = -I include -std=c++11 -O3 $(shell python3-config --cflags) LDFLAGS = $(shell python3-config --ldflags) DEPS = lanms.h $(shell find...
I guess due to some sort of issues caused by data augmentation, some data became unpredictably wrong, and causes the loss of this batch become nan. Seeking which specific training...
@saharudra I can at most achieve 0.7 hmean after modifying the thresholds in `eval.py` on ICDAR 2015 after like 400 epochs. TBH, I don't think this will reproduce the performance...
> Hi bro,I read the code of yours and the src tf version,and I use your code to train but found can get 0.4 hmean on ic2015 test dataset,and I...
@saharudra Sorry, I don't think this will achieve ICDAR 2015 results mentioned on the EAST paper due to the low recall rate and hence low hmean. To be more specific,...
这种情况用整体时间平移可以解决吗？
> Does record represent immutability? Record type is not necessarily asking all members to be immutable. It can still provide useful functions such as a nice ToString format and equal...
In my case, there is a constant 5 lines offset. Both the latest VS Code and Sync Scroll plugin