Abstract
Content-Based Image Retrieval (CBIR) system enables us to access images using only images as queries, instead of keywords. Photorealistic images, and hand-drawn sketch image can be used as a queries as well. Recently, convolutional neural networks (CNNs) are used to discriminate images including sketches. However, the tasks are limited to classifying only one type of images, either photo or sketch images, due to the lack of a large dataset of sketch images and the large difference of their visual characteristics. In this paper, we introduce a simple way to prepare training datasets, which can enable the CNN model to classify both types of images by color transforming photo and illustration images. Through the training experiment, we show that the proposed method contributes to the improvement of classification accuracy.
| Original language | English |
|---|---|
| Title of host publication | Artificial Neural Networks and Machine Learning - 25th International Conference on Artificial Neural Networks, ICANN 2016, Proceedings |
| Publisher | Springer Verlag |
| Pages | 283-290 |
| Number of pages | 8 |
| Volume | 9887 LNCS |
| ISBN (Print) | 9783319447803 |
| DOIs | |
| Publication status | Published - 2016 |
| Event | 25th International Conference on Artificial Neural Networks and Machine Learning, ICANN 2016 - Barcelona, Spain Duration: 2016 Sep 6 → 2016 Sep 9 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 9887 LNCS |
| ISSN (Print) | 03029743 |
| ISSN (Electronic) | 16113349 |
Other
| Other | 25th International Conference on Artificial Neural Networks and Machine Learning, ICANN 2016 |
|---|---|
| Country | Spain |
| City | Barcelona |
| Period | 16/9/6 → 16/9/9 |
Fingerprint
Keywords
- Content based image retrieval
- Hand-drawn sketch
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)
Cite this
Classification of photo and sketch images using convolutional neural networks. / Sasaki, Kazuma; Yamakawa, Madoka; Sekiguchi, Kana; Ogata, Tetsuya.
Artificial Neural Networks and Machine Learning - 25th International Conference on Artificial Neural Networks, ICANN 2016, Proceedings. Vol. 9887 LNCS Springer Verlag, 2016. p. 283-290 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9887 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Classification of photo and sketch images using convolutional neural networks
AU - Sasaki, Kazuma
AU - Yamakawa, Madoka
AU - Sekiguchi, Kana
AU - Ogata, Tetsuya
PY - 2016
Y1 - 2016
N2 - Content-Based Image Retrieval (CBIR) system enables us to access images using only images as queries, instead of keywords. Photorealistic images, and hand-drawn sketch image can be used as a queries as well. Recently, convolutional neural networks (CNNs) are used to discriminate images including sketches. However, the tasks are limited to classifying only one type of images, either photo or sketch images, due to the lack of a large dataset of sketch images and the large difference of their visual characteristics. In this paper, we introduce a simple way to prepare training datasets, which can enable the CNN model to classify both types of images by color transforming photo and illustration images. Through the training experiment, we show that the proposed method contributes to the improvement of classification accuracy.
AB - Content-Based Image Retrieval (CBIR) system enables us to access images using only images as queries, instead of keywords. Photorealistic images, and hand-drawn sketch image can be used as a queries as well. Recently, convolutional neural networks (CNNs) are used to discriminate images including sketches. However, the tasks are limited to classifying only one type of images, either photo or sketch images, due to the lack of a large dataset of sketch images and the large difference of their visual characteristics. In this paper, we introduce a simple way to prepare training datasets, which can enable the CNN model to classify both types of images by color transforming photo and illustration images. Through the training experiment, we show that the proposed method contributes to the improvement of classification accuracy.
KW - Content based image retrieval
KW - Hand-drawn sketch
UR - http://www.scopus.com/inward/record.url?scp=84988373762&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84988373762&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-44781-0_34
DO - 10.1007/978-3-319-44781-0_34
M3 - Conference contribution
AN - SCOPUS:84988373762
SN - 9783319447803
VL - 9887 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 283
EP - 290
BT - Artificial Neural Networks and Machine Learning - 25th International Conference on Artificial Neural Networks, ICANN 2016, Proceedings
PB - Springer Verlag
ER -