Zhihua Huang
Zhihua Huang
I run your code and only the nose and right mouth get the right result. But the rest landmarks' failure rate are 99%. It is wired.
As the title, when I run the command, it do nothing for me. And I set the `$GOPATH`, but there is nothing inside `$GOPATH/src/github.com` . And there is also no...
I use the batch sampler in this repo, but the data fetch speed is quite slow, which is ~250s for 50 batches. The batch size is 256, P=32, K=8, use...
I get same result as paper **except** `challenging set` and `failure rate`. Did I missed something?
hi, in my view, the AffineTransformLayer.py is used to apply the affine/similiarity transform on original training image, but why the affine transform matrix be inversed in lines 59-60?
when I transfer a pytorch model with `torchvision.roi_align` operation, the inference procedure failed with following Error: ``` poptorch.poptorch_core.Error: In poptorch/source/LowerToPopart.cpp:913: 'poptorch_cpp_error': Couldn't find a registered operation for node Error raised...
see title, I wander to know the improvement between normal pooling and pooling with indices.
while i run ./config_gen.py $PROJECT_DIR in the shell, I got below: Running cmake in '/tmp/tmpppC4tz'... $ cmake /home/huangzh/Github/face-alignment-at-3000fps -- The C compiler identification is Clang 3.8.0 -- The CXX compiler...
Develop
use `ctex` package instead of `CJKutf8` package to handle Chinese character. So it no need to use `\zh{}` or `\zhb{}` to wrap the Chinese character.
When I done configuration, it report `error http 404` when I sync the gcal