| KEYWORD |
Virtual validation toolchain for in-cabin sensing
Thesis in external company
keywords AUTOMOTIVE, TESTING, VIRTUAL REALITY
Reference persons RICCARDO COPPOLA
External reference persons The thesis will be conducted at Concept Reply.
Research Groups DAUIN - GR-16 - SOFTWARE ENGINEERING GROUP - SOFTENG
Thesis type COMPANY STAGE
Description Context:
Starting mid-2026, Advanced Driver Distraction Warning (ADDW) Systems become mandatory for new vehicles. Reply is developing a framework for testing and training camera-based Driver and Occupant Monitoring System algorithms in a virtual environment. This reduces real-world testing, offers flexible scenario tuning, and generates Ground Truth data automatically for algorithm testing
Description:
The candidate will join a team focused on validating ML-based Advanced Driver Distraction Warning (ADDW) systems. The primary objective is to create a robust simulation toolchain, utilizing Unreal Engine and Metahuman Animator plugin, to construct databases and ground truth data for training and validating new AI-enhanced Driver Monitoring System (DMS) and Occupant Monitoring System (OMS) functionalities
Objectives:
Face motion tracking and real-time mirroring
Recording human actions and developing a library
Programmatic permutations of recorded scenes (environment, avatar, setup) and API development
Crafting scenario descriptions
Creating a scenario database for specific Driver Monitoring System (DMS) functions
Design and development of a Camera-in-the-Loop-System
Required skills Knowledge of unreal engine
Knowledge of python
Interest in optics
Notes Skills acquired:
Digital human
Game development
DMS/OMS functionalities
Coverage Driven Validation
Scenario-based testing
Camera-in-the-Loop testing
Deadline 26/02/2025
PROPONI LA TUA CANDIDATURA