At Qualcomm I work with object detection and datasets for machine learning.

DescriptionIn what started as the classification department I have worked with detecting pedestrians and vehicles. Initially with classical boosting based classifiers but later on entirely CNN based. I have worked with almost all parts of the product development relating to the object detection chain for two camera generations including requirements, annotation instructions, data collection, data quality assurance, data preprocessing and storage, ML training frameworks, ML verification frameworks, target code, target runtime performance, product performance verification, customer interactions and strategies for future generations. I have also worked with introducing and mentoring new employees and teams which is something I enjoy.

At the "framework" department I primarily worked with scheduling of the vision algorithms on the target hardware and bugs related to it as well as system verification in general.

At Saab I worked with developing an aircraft simulator core and integrating various aircraft subsystems into it.