dukebw / loho Goto Github PK
View Code? Open in Web Editor NEWDemo code for "LOHO: Latent Optimization of Hairstyles via Orthogonalization".
Demo code for "LOHO: Latent Optimization of Hairstyles via Orthogonalization".
Thank you for sharing this work!
I tried to run LOHO with my own images, using Graphonomy to extract masks as metioned in the paper, but I dont know how to get softmasks for soft-blending.
I have been able to implement and understand the logic behind the alignment code of segmentation masks but was not able to understand how did you arrive at the values for categorizing the samples into easy, medium and hard, particularly the PD range values i.e how did you normalize the distance values.
Could you please share the code or some details about how you computed these values?
Hi Team,
Thankyou for your work. Could you please let me know whether the hair work on full body images or not? I have attached file for reference. I need the output to be same as full body but not portrait.
Is that possible anyway to get full body as output without face modification?
Thank you for your hard work!
I'd like to know how to train the logo.
I want to run it with the data I have instead of the sample data provided, what should I do?
How can I make a mask of the data I have?
Thanks for your excellent work!
I downloaded your code and make it run like what your README says, but I found it takes about 16mins to get a result. Is this normal or did I miss some important steps?
My gpu is one GTX 2080Ti.
I am encountering the following error although I have correctly provided paths of every image but still I am getting this error :
FileNotFoundError: [Errno 2] No such file or directory: '/content/LOHO/data/images/00761.png'
I am working on Google coolab and running the following code:
!python /content/LOHO/loho.py --image1 /content/LOHO/data/images/00761.jpg --image2 /content/LOHO/data/images/02602.jpg --image3 /content/LOHO/data/images/00018.jpg
Kindly Help!
Thanks for your work!
It'd be great if you could provide guidelines for training LOHO!
It`s an impressive work.I want test my face images,but I do not have crresponding softmasks and inpaintbackgroud,can you share how to get that.Thank you so much.
Thank you for your great work, can you please provide me with the code for generating inpainted backgrounds, masks, softmasks? Thank you very much!
i notice that you don't have the label of the train img, so how to calculate the ssim and psnr. Thanks a lot~
`Setting up Perceptual loss...
Loading model from: /content/LOHO/networks/lpips/weights/v0.1/vgg.pth
...[net-lin [vgg]] initialized
...Done
Setting up Perceptual loss...
Loading model from: /content/LOHO/networks/lpips/weights/v0.1/vgg.pth
...[net-lin [vgg]] initialized
...Done
Constructing DeepLabv3+ model...
Number of classes: 20
Output stride: 16
Number of Input Channels: 3
/content/drive/MyDrive/inference.pth
Traceback (most recent call last):
File "/content/LOHO/loho.py", line 180, in <module>
state_dict = torch.load(graphonomy_model_path)
File "/usr/local/lib/python3.7/site-packages/torch/serialization.py", line 595, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "/usr/local/lib/python3.7/site-packages/torch/serialization.py", line 781, in _legacy_load
deserialized_objects[key]._set_from_file(f, offset, f_should_read_directly)
RuntimeError: unexpected EOF, expected 1362420 more bytes. The file might be corrupted.`
After Executing loho.py I am getting this error.
Hi,
Really thank you for sharing you works.
By the way, I suppose that networks/lpips/init.py should be fixed. May need to fix line 6.
from skimage.metrics import structural_similarity as compare_ssim
instead of
from skimage.measure import compare_ssim
Thanks again.
Cheers,
Dale
It's an impressive project. I'd love to try out my face and would appreciate it if you could tell me how to create softmasks and inpaintbackground.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ 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. ๐๐๐
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 โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.