izpis_h1_title_alt

Improving convolutional neural networks performance for image classification using test time augmentation : a case study using MURA dataset
ID Kandel, Ibrahem (Author), ID Castelli, Mauro (Author)

.pdfPDF - Presentation file, Download (1,82 MB)
MD5: 58E488E1D8AB9FA76521D568236D9BD4
URLURL - Source URL, Visit https://link.springer.com/article/10.1007%2Fs13755-021-00163-7 This link opens in a new window

Abstract
Bone fractures are one of the main causes to visit the emergency room (ER); the primary method to detect bone fractures is using X-Ray images. X-Ray images require an experienced radiologist to classify them; however, an experienced radiologist is not always available in the ER. An accurate automatic X-Ray image classifier in the ER can help reduce error rates by providing an instant second opinion to the emergency doctor. Deep learning is an emerging trend in artificial intelligence, where an automatic classifier can be trained to classify musculoskeletal images. Image augmentations techniques have proven their usefulness in increasing the deep learning model's performance. Usually, in the image classification domain, the augmentation techniques are used during training the network and not during the testing phase. Test time augmentation (TTA) can increase the model prediction by providing, with a negligible computational cost, several transformations for the same image. In this paper, we investigated the effect of TTA on image classification performance on the MURA dataset. Nine different augmentation techniques were evaluated to determine their performance compared to predictions without TTA. Two ensemble techniques were assessed as well, the majority vote and the average vote. Based on our results, TTA increased classification performance significantly, especially for models with a low score.

Language:English
Keywords:image classification, convolutional neural networks, transfer learning, test time 29 augmentation, deep learning, ensemble learning
Work type:Article
Typology:1.01 - Original Scientific Article
Organization:EF - School of Economics and Business
Publication status:Published
Publication version:Author Accepted Manuscript
Year:2021
Number of pages:22 str.
Numbering:Vol. 9, art 33
PID:20.500.12556/RUL-129209 This link opens in a new window
UDC:659.2:004
ISSN on article:2047-2501
DOI:10.1007/s13755-021-00163-7 This link opens in a new window
COBISS.SI-ID:72230659 This link opens in a new window
Publication date in RUL:30.08.2021
Views:723
Downloads:173
Metadata:XML RDF-CHPDL DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Record is a part of a journal

Title:Health information science and systems
Shortened title:Health inf. sci. syst.
Publisher:Springer
ISSN:2047-2501
COBISS.SI-ID:523110681 This link opens in a new window

Projects

Funder:FCT - Fundação para a Ciência e a Tecnologia, I.P.
Project number:DSAIPA/DS/0022/2018
Name:GADgET

Funder:ARRS - Agencija za raziskovalno dejavnost Republike Slovenije
Funding programme:Raziskovalni program
Project number:P5-0410
Name:Digitalizacija kot gonilo trajnostnega razvoja posameznika, organizacij in družbe

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back