juefeix / pnn.pytorch Goto Github PK
View Code? Open in Web Editor NEWPyTorch implementation of CVPR'18 - Perturbative Neural Networks http://xujuefei.com/pnn.html
License: MIT License
PyTorch implementation of CVPR'18 - Perturbative Neural Networks http://xujuefei.com/pnn.html
License: MIT License
I read in the paper
When the input image comes in, it will be perturbed with a set of pre-defined random additive noise masks
Does this code always generate the same noise for each run for train and for evaluate?
if self.noise.numel() == 0:
self.noise.resize_(x.data[0].shape).uniform_()
self.noise = (2 * self.noise - 1) * self.level
Firstly, Thank you for the amazing work. The inference speed is my main concern for now. Can you please compare it with CNN, especially when m/p=256?
Dear @juefeix ,
I could not figure out how to control fan-out ratio using your code. Would you please to make this clear for me?
The latest visdom version gives an error.
AttributeError: 'Visdom' object has no attribute 'updateTrace'
When I installed visdom version of 0.17, no error occurred.
Thanks for releasing your code! Any chance you might be able to release some of the trained models reported in your paper, especially the model trained on ImageNet?
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.