COPYRIGHT
This dataset can be used for any non-commercial purpose, e.g., research or education, given that the following paper is adequately cited:
Alexander Freytag and Erik Rodner and Marcel Simon and Alexander Loos and Hjalmar Kühl and Joachim Denzler: "Chimpanzee Faces in the Wild: Log-Euclidean CNNs for Predicting Identities and Attributes of Primates", German Conference on Pattern Recognition (GCPR), 2016 .
PROVIDED DATA
Images
We provide two datasets of cropped chimpanzee faces: C-Zoo and C-Tai. For details of both datasets, please refer to the corresponding paper (e.g., available at this pdf link ).
Labels
In addition to the pure images, the dataset comes with annotations of identity, gender, age, and age group. All annotations have been provided by experts. More details about statistics of data are shown in the supplementary material (e.g., available at this pdf link ).
Train/Test Splits
We provide five splits intro train and test which have been used to produce the results in Table 1 of the GCPR'16 paper. The demo script (see below) shows how to access data accordingly.
HOW TO ACCESS
Access via Matlab (recommended)
We provided a tiny matlab script which loads one of the datasets, requests the splits, and loops through the images. Depending on your needs, you might want to use this as a starting point for your own evaluations.
cd demo_access_data/
demo_chimp_loading
Access via Textfiles (not recommended)
For persons without access to Matlab, we additionally provide the entire set of attributes in a separate text file for each dataset. To read all data, go to the dataset directory:
cd datasets_cropped_chimpanzee_faces/data_CZoo/
and look into the provided annotations file:
vim annotations_czoo.txt
Layout of the annotations is as follows.
- all information for a single image in one line
- fields are separated by white space
- linebreak is done by \n
- Keypoints are given as (x-coordinate y-coordinate) pairs
Order of fields is
- Filename (string)
- Age (float, NaN if unavailable)
- Age_Group (string, Unknown if unavailable)
- Gender (string, Unknown if unavailable)
- Keypoint RightEye (int, Inf if unavailable)
- Keypoint LeftEye (int, Inf if unavailable)
- Keypoint MouthCenter (int, Inf if unavailable)
- Keypoint LeftEarlobe (int, Inf if unavailable)
- Keypoint RightEarlobe (int, Inf if unavailable)
ACKNOWLEDGEMENTS
The authors thank Dr. Tobias Deschner for providing the images which were used to build the C-Tai dataset, Laura Aporius and Karin Bahrke for collecting and annoting the images which were used to build the C-Zoo dataset, and the Zoo Leipzig for providing permission for image collection. The images used for creating the C-Zoo dataset were collected as part of the SAISBECO project funded by the Pact for Research and Innovation between the Max Planck Society and the Fraunhofer-Gesellschaft.