This dataset was generated from Multi-PIE face dataset [1, 2]. In this repository, we just published image name from Multi-PIE face dataset and detected ear coordinates. You can use face images and corresponding coordinates to obtain ear images.
This dataset is the extended version of Multi-PIE ear dataset [3].
This dataset has been explained and has been used in our paper. Please cite it, if it is helpful for you.
@inproceedings{yaman2019multimodal,
title={Multimodal Age and Gender Classification Using Ear and Profile Face Images},
author={Yaman, Dogucan and Irem Eyiokur, Fevziye and Kemal Ekenel, Hazim},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops},
pages={0--0},
year={2019}
}
CSV file contains following headers:
| Name | Description |
|---|---|
| Image Name | Original image name in MultiPie Face dataset |
| ID | Original ID of the corresponding subject in MultiPie Face dataset |
| Label | Subject label for person identification |
| Ear Side | Indicate whether the corresponding ear is left or right one |
| x | x coordinate of the starting point of the bounding box |
| y | y coordinate of the starting point of the bounding box |
| w | Width of the bounding box |
| h | Height of the bounding box |
Note: Ear detection script, train, and test files will be uploaded.
- [1] Gross, R., Matthews, I., Cohn, J.F., et al.: ‘Multi-PIE’. IEEE Int. Conf. on Automatic Face and Gesture Recognition (FG), 2008.
- [2] Gross, R., Matthews, I., Cohn, J.F., et al.: ‘Multi-PIE’. Image and Vision Computing, 2010, pp. 807–813.
- [3] Eyiokur, F.I., Yaman, D., and Ekenel, H.K.: "Domain adaptation for ear recognition using deep convolutional neural networks.", IET Biometrics 7.3 (2017): 199-206.