How to reproduce evaluation results

As part of the master thesis, there were some crucial results generated that evaluated not only the object detector models but also the other parts of the pipeline as well. In order to encourage future work, it is important to establish baselines in the form of evaluation results achieved using current approaches. Keeping that in mind, special care has been taken so that the reproduction of the evaluation results is easy. This is done so using a makefile that accompanies the code.

Tested On

The code has been tested to run on:

• macOS 12.4 Monterey on Intel Macbook Pro 2019
• Windows 11 on Intel Core i5 using Docker