ningshuliang / picture Goto Github PK
View Code? Open in Web Editor NEWOfficial code for paper "PICTURE: PhotorealistIC virtual Try-on from UnconstRained dEsigns"
License: MIT License
Official code for paper "PICTURE: PhotorealistIC virtual Try-on from UnconstRained dEsigns"
License: MIT License
Looks very good
When will the code be released?
I succeed to inference your work with text clip in stage1, but I saw blurry results when using img clip
I tried below code
from ldm.modules.encoders.modules import ClipImageProjector
from torchvision import transforms
version="./../pretrain_models/clip-vit-large-patch14"
clip_model2 = ClipImageProjector(version=version).to(device)
tform = transforms.ToTensor()
# text = tokenizer(text_description,truncation=True, max_length=77, return_length=True,
# return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
# text_features = clip_model(text["input_ids"].cuda(non_blocking=True))
# text_features = text_features.last_hidden_state # torch.Size([1, 77, 768])
garment_condition_path = os.path.join("./Sample_data/Cloth_White_Background", file_name[0])
garment_condition = tform(Image.open(garment_condition_path).convert("RGB"))
garment_condition = garment_condition * 2. - 1.
garment_condition = clip_model2.preprocess(garment_condition.unsqueeze(0)) # got similar results with / without preprocessing
text_features = clip_model2(garment_condition.cuda(non_blocking=True))
c = [concat_feature,text_features]
sampler.sample(S=opt.ddim_steps,
conditioning=c,
...)
Could you please help me to use img clip?
Google link show permission error.
I'm trying to make parsing imgs similar to what you provided,
but after the segmentation generation process from VITON-HD, I got imgs similar to below figure.
Could you please provide just a little bit of code snippets, about color and labels?
For example,
labels = { # from https://github.com/shadow2496/VITON-HD
0: ['background', [0]],
1: ['paste', [2, 4, 7, 8, 9, 10, 11]],
2: ['upper', [3]],
3: ['hair', [1]],
4: ['left_arm', [5]],
5: ['right_arm', [6]],
6: ['noise', [12]]
}
labels_to_colours = [(0,0,0), ...]
Thank you for your contribution to community with this repo ! I can't reach pre trained weights when i click pre trained weights link. Will you share them?
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.