Abstract
Imbalanced dataset problem may occur when the number of instances of a certain class is much lower than others, resulting in a drop in the classification result of minority class. We propose the method of generating images from 3D modeling by some softwares to get enough images of minority class and supplement the dataset to re-balance it. Several deep networks are trained on these datasets. The experiment results are evaluated by F-measure and show that when the images are generated by enough models, the classification performance can be obviously improved.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the 36th Chinese Control Conference, CCC 2017 |
| Editors | Tao Liu, Qianchuan Zhao |
| Publisher | IEEE Computer Society |
| Pages | 10930-10935 |
| Number of pages | 6 |
| ISBN (Electronic) | 9789881563934 |
| DOIs | |
| Publication status | Published - 7 Sept 2017 |
| Event | 36th Chinese Control Conference, CCC 2017 - Dalian, China Duration: 26 Jul 2017 → 28 Jul 2017 |
Publication series
| Name | Chinese Control Conference, CCC |
|---|---|
| ISSN (Print) | 1934-1768 |
| ISSN (Electronic) | 2161-2927 |
Conference
| Conference | 36th Chinese Control Conference, CCC 2017 |
|---|---|
| Country/Territory | China |
| City | Dalian |
| Period | 26/07/17 → 28/07/17 |
Keywords
- 3D modeling
- deep learning
- image recognition
Fingerprint
Dive into the research topics of 'Generating images for imbalanced dataset problem'. Together they form a unique fingerprint.Cite this
Qin, Y., Chen, W., & Chen, J. (2017). Generating images for imbalanced dataset problem. In T. Liu, & Q. Zhao (Eds.), Proceedings of the 36th Chinese Control Conference, CCC 2017 (pp. 10930-10935). Article 8029100 (Chinese Control Conference, CCC). IEEE Computer Society. https://doi.org/10.23919/ChiCC.2017.8029100