materight / repnet-pytorch Goto Github PK
View Code? Open in Web Editor NEWA PyTorch port with pre-trained weights of RepNet, from "Counting Out Time: Class Agnostic Video Repetition Counting in the Wild".
A PyTorch port with pre-trained weights of RepNet, from "Counting Out Time: Class Agnostic Video Repetition Counting in the Wild".
Hi,
Thanks for the implementation.
I was just wondering if the horizontal and vertical dark lines (like a grid) on the embedding plots are correct?
They don't seem to be present on any of the official implementation examples https://sites.google.com/view/repnet
I've had a look through the code but can't really work out why they're there, other than maybe some issue with the frame embeddings or frame sampling?
Cheers, Matthew.
Im really confused on this one. The original repo, would let you feed as many frames as you want, and get the score for rep of each frame.
When using this repo, it doesnt provide the original functionally that allows for this, it also takes a 10sec video, 'says' its only checking 1 second for some reason, and then it outputs a 6 sec video ? Seems all over the place.
Would it be possible to get the original functionally back ? This acts more like a demo or showcase that workable code. Or am i missing something.
If someone wants to try to run the official google colab code from https://sites.google.com/view/repnet, I found this error : ValueError: order_filterND not available for this type. If you ignore the median_filter then you can test some videos. Another issue that I found is in the function create_count_video, parameter b seems to be depracated since matplotlib 3.5, accoding to: matplotlib/matplotlib#25267
I hope this could be useful to someone.
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.