caenorst / yenet-pytorch Goto Github PK
View Code? Open in Web Editor NEWPytorch implementation of "Deep Learning Hierarchical Representations for Image Steganalysis"
Pytorch implementation of "Deep Learning Hierarchical Representations for Image Steganalysis"
when I run the program,I meet blow error:
RuntimeError :Expected 4-dimensional input for 4-dimensional weight [30,1,5,5],but got 5-
dimensional input of size [16,2,1,256,256] instead
I don't know why?please help me!
Dear author
I'm using the BOSSbase_1.01 dataset, WOW algorithm 0.4bpp
4000 train, 1000 validation.
I train the model but I get valid_accuracy=0.5
could you help me what went wrong?
in train_loader, pair_constraint=not(args.use_batch_norm), Why set pair_constraint as False when use batch normalization?
In your implementation, the image don't be normalized to [0.0,1.0]?
If the pair_constraint = false, the code will use DatasetNoPair. Like DatasetPair, the DatasetNoPair also can call transform (RandomRot, RandomFlip). In paired mode, the cover and stego images will be both rotated or flipped using the same setting.
However, when in no-pair mode, it can rotate or flip the cover image only, the corresponding stego image may not be rotated or flipped. Because the DatasetNoPair only load one image once.
Therefore, if set pair_constraint = false, the only transforms.Compose([
utils.ToTensor()
]) can be used?
the dataloaditer is out of date in pytorch1.0, how should I rewrite it or do something to instead? I use the dataloadFolder instead without pair, but the network seems didn't work.
Dear author:
Thanks for your implementation about YeNet. However, I want to know whether your source code don't include the part of incorporating the knowledge of selection-channel? In other words. you have implemented TLU-CNN but not SCA-TLU-CNN. Is it right? Because I have found the embedding probability map is just used to generate the stego image (if embedding_otf=True). Would you mind sharing the updating code about implementing SCA-TLU-CNN?
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.