| IEICE Technical Committee Submission System Conference Paper's Information |
Online Proceedings [Sign in] Tech. Rep. Archives |
| Go Top Page | Go Previous | [Japanese] / [English] |
| Paper Abstract and Keywords | ||
| Presentation | 2022-09-14 16:15 Data Augmentation with Style Transfer for Fossil Image Segmentation Akihiro Waza (Osaka Metropolitan Univ.), Yuya Inamura (Osaka Prefecture Univ.), Katsufumi Inoue, Michifumi Yoshioka, Toshihiro Yamada (Osaka Metropolitan Univ.)  PRMU2022-17 |
|
| Abstract | (in Japanese) | (See Japanese page) |
| (in English) | Fossils are extremely important materials in evolutionary biology and earth science. However, it is necessary to have specialized knowledge, experience, time, and effort in order to discover and excavate the fossils. Therefore, there is a need for a support system that automatically detecting fossils. In this research, as the first step of the support system, we developed a semantic segmentation method to recognize the fossil parts exposed from rocks. In general, semantic segmentation by deep learning requires a large number of images. However, the number of fossil images is very small, making it difficult to train deep learning models. In this research, we approach this problem by data augmentation. Specifically, we used images of plant specimens and sedimentary rocks since they are easily available. In addition, we generated images that closely resemble real fossil images from them by using a style transfer method. Then we used this images for data augmentation. Experimental results showed that DeepLabv3+, a representative deep learning model for semantic segmentation, improved the F-score by about 5% compared to that before data augmentation. | |
| Keyword | (in Japanese) | (See Japanese page) |
| (in English) | Fossil Images / Style Transfer / Data Augmentation / Deep Learning / Semantic Segmentation / / / | |
| Reference Info. | IEICE Tech. Rep., vol. 122, no. 181, PRMU2022-17, pp. 43-48, Sept. 2022. | |
| Paper # | PRMU2022-17 | |
| Date of Issue | 2022-09-07 (PRMU) | |
| ISSN | Online edition: ISSN 2432-6380 | |
| Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) | |
| Download PDF | PRMU2022-17 |
|
| Conference Information | |
| Committee | PRMU |
| Conference Date | 2022-09-14 - 2022-09-15 |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Deep generative model |
| Paper Information | |
| Registration To | PRMU |
| Conference Code | 2022-09-PRMU |
| Language | Japanese |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Data Augmentation with Style Transfer for Fossil Image Segmentation |
| Sub Title (in English) | |
| Keyword(1) | Fossil Images |
| Keyword(2) | Style Transfer |
| Keyword(3) | Data Augmentation |
| Keyword(4) | Deep Learning |
| Keyword(5) | Semantic Segmentation |
| Keyword(6) | |
| Keyword(7) | |
| Keyword(8) | |
| 1st Author's Name | Akihiro Waza |
| 1st Author's Affiliation | Osaka Metropolitan University (Osaka Metropolitan Univ.) |
| 2nd Author's Name | Yuya Inamura |
| 2nd Author's Affiliation | Osaka Prefecture University (Osaka Prefecture Univ.) |
| 3rd Author's Name | Katsufumi Inoue |
| 3rd Author's Affiliation | Osaka Metropolitan University (Osaka Metropolitan Univ.) |
| 4th Author's Name | Michifumi Yoshioka |
| 4th Author's Affiliation | Osaka Metropolitan University (Osaka Metropolitan Univ.) |
| 5th Author's Name | Toshihiro Yamada |
| 5th Author's Affiliation | Osaka Metropolitan University (Osaka Metropolitan Univ.) |
| 6th Author's Name | |
| 6th Author's Affiliation | () |
| 7th Author's Name | |
| 7th Author's Affiliation | () |
| 8th Author's Name | |
| 8th Author's Affiliation | () |
| 9th Author's Name | |
| 9th Author's Affiliation | () |
| 10th Author's Name | |
| 10th Author's Affiliation | () |
| 11th Author's Name | |
| 11th Author's Affiliation | () |
| 12th Author's Name | |
| 12th Author's Affiliation | () |
| 13th Author's Name | |
| 13th Author's Affiliation | () |
| 14th Author's Name | |
| 14th Author's Affiliation | () |
| 15th Author's Name | |
| 15th Author's Affiliation | () |
| 16th Author's Name | |
| 16th Author's Affiliation | () |
| 17th Author's Name | |
| 17th Author's Affiliation | () |
| 18th Author's Name | |
| 18th Author's Affiliation | () |
| Speaker | Author-1 |
| Date Time | 2022-09-14 16:15:00 |
| Presentation Time | 15 minutes |
| Registration for | PRMU |
| Paper # | PRMU2022-17 |
| Volume (vol) | vol.122 |
| Number (no) | no.181 |
| Page | pp.43-48 |
| #Pages | 6 |
| Date of Issue | 2022-09-07 (PRMU) |
[Return to Top Page]