.
Datasets & Tools
For each task comprising the HEP-2 CONTEST 2016, the participants receive a software tool for computing the actual performance indices on the training set for each task, in order to check and tune their algorithms.
| TASK | Type | Description | Version | Size | Download |
| 1 | Dataset | The file contains the train set for the Task 1. This dataset was already used during the ICIP 2013 and ICPR 2014 contests. | 1.0 | 28 MB |  |
| 2 & 4 | Dataset | The file contains the train set for the Tasks 2 and 4. This dataset was already used during the ICPR 2014 contests but only for the Task 2, while it is used for the first time for the Task 4. | 1.0 | 482 MB | |
| 3 | Dataset |
The file contains the train set for the Task 3. This dataset is newly introduced for this competition. |
1.0 | 8 MB | |
| 1 & 2 | Software Tools | Python scripts that can be used to run the application on all the images of the dataset and to compare the results with the ground truth. Note that these tools can be used to evaluate the algorithm for both Tasks 1 and 2. | 1.0 | 1M | |
| 3 | Software Tools | Python scripts that can be used to run the application on all the images of the dataset for the Task 3 and to compare the results with the ground truth. | 1.0 | 1M | |
| 4 | Software Tools | Python scripts that can be used to run the application on all the images of the dataset for the Task 4 and to compare the results with the ground truth. | 1.0 | 1M | |
| 1-4 | Documentation | Readme file with instructions about the use of the software tools for performance evaluation | 1.0 | 1M | |