Results on the CIFAR-100 dataset.
dataset
posted on 2025-02-21, 18:35 authored by Hairui Wang, Mengjie Dong, Guifu Zhu, Ya LiWe tested the results with different values of: 1, 0.5, and 0.25.
History
Usage metrics
Categories
Keywords
wasted computational resourcestypically employed independentlytarget classes obtainedtarget classes learnedobject detection tasksclassifier &# 8217integrating distillation methodstwo methodstwo componentssingle methodrespective advantagesredundant informationonline distillationoffline distillationnew perspectiveinformation conveyedincreased complexityimagenet datasetsimage classificationhighly similargreat potentialdistillation researchdckd ),correct knowledgecompromise solutionalready acquiredallows relational
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC