Qualcomm FY25 Intern - Thesis-Approaching end end-to-end 

Sweden, Linköping 

907192117

Arriver Software AB

Job Area:

Interns Group, Interns Group > Interim Intern

Approaching the end of end-to-end high beam assist

The current trend in both industry and academia suggests end-to-end (E2E) machine learning solutions for automated driving systems. An E2E system is globally optimized towards the system’s end goal, in contrast to a modular system in which each module is locally optimized towards some intermediate proxy goal.

Other possible benefits with an E2E solution are that less annotation would be needed, that the development and maintenance of the system would be easier, and that the computational cost would be lower. However, some possible drawbacks with an E2E solution are that it would mean a system more “black box” in nature – with implications for safety, interpretability and debugging – and that it would be harder to train.

One could also consider hybrid solutions, which are less modular but not fully end-to-end.

This master thesis will investigate approaching an end-to-end solution for the automated driving sub-system of high beam assist.

Learning Objectives:

Improved knowledge of the benefits, drawbacks and feasibility of approaching an end-to-end machine learning solution for an automated driving sub-system.

Student Background:

Master studies, experience with Python and Machine Learning (with Deep Learning and Computer Vision being particularly meritorious).

Thesis work suitable for 1-2 students.

*References to a particular number of years experience are for indicative purposes only. Applications from candidates with equivalent experience will be considered, provided that the candidate can demonstrate an ability to fulfill the principal duties of the role and possesses the required competencies.

Qualcomm expects its employees to abide by all applicable policies and procedures, including but not limited to security and other requirements regarding protection of Company confidential information and other confidential and/or proprietary information, to the extent those requirements are permissible under applicable law.