Transfer learning of deep learning vision models
Title: Transfer learning of deep learning vision models
SNIC Project: Berzelius-2022-159
Project Type: LiU Berzelius
Principal Investigator: Mohammad Moein Sorkhei <sorkhei@kth.se>
Affiliation: Kungliga Tekniska högskolan
Duration: 2022-08-15 – 2023-03-01
Classification: 10201
Keywords:

Abstract

Transfer learning for vision models is a concept where the goal is to use the knowledge acquired by a network in one domain/dataset in another domain/dataset possibly with limited data. The goal of this project is to investigate the impact of transfer learning techniques in image classification, essentially investigating how much such a techniques helps improve the performance of deep learning image classification models in various image classification benchmarks and settings. This is done by training deep learning models for the task of image classification using various datasets.