szq0214/FKD
Official code for our ECCV'22 paper "A Fast Knowledge Distillation Framework for Visual Recognition"
This framework helps machine learning engineers and researchers to improve the performance of smaller, faster image recognition models. By leveraging "knowledge distillation" with pre-generated soft labels, it allows you to train a compact model using insights from a larger, more powerful model. The result is a more efficient visual recognition system that maintains high accuracy.
191 stars. No commits in the last 6 months.
Use this if you need to deploy accurate visual recognition models on resource-constrained devices or in environments where fast inference is critical.
Not ideal if you are developing models from scratch without a larger, pre-trained 'teacher' model, or if computational efficiency is not a primary concern.
Stars
191
Forks
30
Language
Python
License
MIT
Category
Last pushed
Apr 29, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/szq0214/FKD"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Guang000/Awesome-Dataset-Distillation
A curated list of awesome papers on dataset distillation and related applications.
dkozlov/awesome-knowledge-distillation
Awesome Knowledge Distillation
SforAiDl/KD_Lib
A Pytorch Knowledge Distillation library for benchmarking and extending works in the domains of...
SakurajimaMaiii/ProtoKD
[ICASSP 2023] Prototype Knowledge Distillation for Medical Segmentation with Missing Modality
HikariTJU/LD
Localization Distillation for Object Detection (CVPR 2022, TPAMI 2023)