Few-shot learning for image classification
Thesis event information
Date and time of the thesis defence
Place of the thesis defence
L10, Linnanmaa
Topic of the dissertation
Few-shot learning for image classification
Doctoral candidate
Master of Science Yawen Cui
Faculty and unit
University of Oulu Graduate School, Faculty of Information Technology and Electrical Engineering, Center for Machine Vision and Signal Analysis
Subject of study
Computer Science and Engineering
Opponent
Professor Karen Eguiazarian, Tampere University
Custos
Professor Matti Pietikäinen, University of Oulu
Few-Shot Learning for Image Classification
This thesis contributes to the research on Few-Shot Learning (FSL) for image classification from two aspects:
1) Unsupervised FSL (UFSL): How to transfer prior knowledge learned from a fully unlabeled auxiliary dataset to novel tasks with a few examples?
2) Few-Shot Continual Learning (FSCL): How to continually learn FSL tasks?
1) Unsupervised FSL (UFSL): How to transfer prior knowledge learned from a fully unlabeled auxiliary dataset to novel tasks with a few examples?
2) Few-Shot Continual Learning (FSCL): How to continually learn FSL tasks?
Last updated: 23.1.2024