csxmli2016 / asffnet512 Goto Github PK
View Code? Open in Web Editor NEWThis is a totally reimplemented version of our ASFFNet (CVPR2020) on 512*512 images.
This is a totally reimplemented version of our ASFFNet (CVPR2020) on 512*512 images.
I found that I could enable CUDA support by installing the pytorch version for RTX 3.x, but the code threw an error about OMP: Error #15. The hacky workaround I found was adding:
`os.environ['KMP_DUPLICATE_LIB_OK']='True'
after line 9: import os
in main_test.py
with that small change, and pytorch with CUDA support, CUDA works
You almost certainly have a better solution than what I found, but I thought I'd pass along it anyway, for what it's worth.
I realize this project is very new, but publishing your requirements.txt would make testing the code much easier.
Can you provide the training code?
(net) C:\Users\leolo\ASFFNet512>python main_test.py -s ./TestExamples/TestLists.txt -d ./TestExamples/TestResults
Using device : cpu
Model params : 23.63M
Restoring i1.png
Traceback (most recent call last):
File "main_test.py", line 159, in <module>
optimal_ref_landmarks, optimal_ref_path = optimal_reference_selection(lq_landmarks, land_path)
File "main_test.py", line 80, in optimal_reference_selection
tmp = [np.float(i) for i in line.split(' ') if i != '\n']
File "main_test.py", line 80, in <listcomp>
tmp = [np.float(i) for i in line.split(' ') if i != '\n']
File "C:\Users\leolo\anaconda3\envs\net\lib\site-packages\numpy\__init__.py", line 305, in __getattr__
raise AttributeError(__former_attrs__[attr])
AttributeError: module 'numpy' has no attribute 'float'.
`np.float` was a deprecated alias for the builtin `float`. To avoid this error in existing code, use `float` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.float64` here.
The aliases was originally deprecated in NumPy 1.20; for more details and guidance see the original release note at:
https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
输入需要有HQ参考图像,还是不符合通用场景,大多数人脸都无法获取到参考HQ图像
In one of your examples with Obama and Rami Malek you show eye color being transferred over from Rami to Obama....can the landmark list be filtered to only transfer eyes while leaving the rest untouched? #
Cool work!
Are there any volunteer will make the code the Colab-notebook?
I would like to try this code with V100 on colab.
getting the following error
(net-cuda) C:\Users\leolo\ASFFNet512\TestExamples>python AlignAsFFHQ.py -s ./ObamaRaw -d ./HQReferences/Obama
Traceback (most recent call last):
File "AlignAsFFHQ.py", line 108, in <module>
landmarks_detector = face_alignment.FaceAlignment(face_alignment.LandmarksType._3D, flip_input=False)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\face_alignment\api.py", line 77, in __init__
self.face_detector = face_detector_module.FaceDetector(device=device, verbose=verbose, **face_detector_kwargs)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\face_alignment\detection\sfd\sfd_detector.py", line 31, in __init__
self.face_detector.to(device)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\torch\nn\modules\module.py", line 1145, in to
return self._apply(convert)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
module._apply(fn)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\torch\nn\modules\module.py", line 820, in _apply
param_applied = fn(param)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\torch\nn\modules\module.py", line 1143, in convert
return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
File "C:\Users\leolo\anaconda3\envs\net-cuda\lib\site-packages\torch\cuda\__init__.py", line 239, in _lazy_init
raise AssertionError("Torch not compiled with CUDA enabled")
AssertionError: Torch not compiled with CUDA enabled
I tried creating and enviroment with this and running
conda create --name net python=3.8
and another attempt with
conda create --name net-cuda python=3.8 cudatoolkit=11.3
whats the best way to set up this enviroment on windows, same thing happened with DMDNet but was able to use cpu with that however now would like to be using reference images so came here
Where can I get VGGFace2 and CASIA-WebFace Dataset?
I couldn't get it even if I went into the homepage.
Thx very much!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.