dkozlov / awesome-knowledge-distillation Goto Github PK
View Code? Open in Web Editor NEWAwesome Knowledge Distillation
License: Apache License 2.0
Awesome Knowledge Distillation
License: Apache License 2.0
Hi @dkozlov
I notice that there is no paper in 2024 listed yet. Could you please add a relevant paper from CVPR 2024 Highlight. The information is given below:
Title: Logit Standardization in Knowledge Distillation
Paper: https://arxiv.org/abs/2403.01427
Github: https://github.com/sunshangquan/logit-standardization-KD
Supplements: https://sunsean21.github.io/resources/cvpr2024_supp.pdf
The paper discusses the possibility of assigning temperatures distinctly between teacher/student and dynamically across samples. It then proposes a weighted Z-score logit standardization as a plug-and-play preprocess, capable of boosting the existing logit-based KD methods.
Thank you for your attention.
Thanks for creating this repository and gathering lots of useful papers and materials. I think you can add the following paper to this list:
Also, there is a good implementation of this paper in PyTorch.
Some papers are far away from "awesome". For example, the paper <<Model Distillation with Knowledge Transfer from Face Classification to Alignment and Verification>> was rejected by ICLR2018. Maybe the paper is still unpublished.
I hope that the owner of this repo could be more strict with the quality of the listed papers, and append more information about the publication status of the papers. If some paper got rejected, it's likely that it's not an "awesome" paper
Knowledge Adaptation for Efficient Semantic Segmentation
https://arxiv.org/abs/1903.04688
Structured Knowledge Distillation for Semantic Segmentation
https://arxiv.org/abs/1903.04197
A large scale Awesome KD repository containing more papers:
https://github.com/FLHonker/Awesome-Knowledge-Distillation
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.