This code contains the scripts utilized by VOGUES, intended to provide a greater understanding for how it works.
The code is split into 3 parts: pose estimation (for surveillance/human evaluation), vehicle pose estimation (for autonomous driving/vehicle evaluation), and validation (for the LSTM validator). The L2 attack implementation for adaptive attacks is also included, as a separate script.