University of Twente Student Theses

Login

Constructing a Convolutional Neural Network for Semantic Segmentation of Skin Lesions using a Small Dataset

Kivits, M.P.W. (2018) Constructing a Convolutional Neural Network for Semantic Segmentation of Skin Lesions using a Small Dataset.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Embargo date:1 November 2068
Abstract:Early detection of melanoma is difficult for the human eye but a crucial step towards reducing its death rate. Computerized detection of these melanoma and other skin lesions is necessary. The research question in this paper is "How to segment skin lesion images using a neural network with low available data? This question is divided into three sub-questions regarding the best-performing network structure, training data and training method. The literature states that U-net CNN structures have excellent performances on the segmentation task, more training data increases network performance and utilizing transfer learning increases network generalization performance. Two experiments are conducted. The first experiment trains a network on data sets of different size. The second experiment proposes twelve network structures and trains them on the same data set. The experimental results support the findings in the literature. The best performing skin lesion segmentation network has a fully convolutional structure with a skip architecture and an encoder depth of either one or two. Weights of this network should be initialized using transfer learning from the pre-trained VGG16 network. Training data should be cropped to reduce complexity and augmented during training to reduce the likelihood of overfitting.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:53 electrotechnology, 54 computer science
Programme:Electrical Engineering BSc (56953)
Link to this item:http://purl.utwente.nl/essays/77121
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page