dev-ops

Bosch Group

<div><section><div><p></p><h2>Company Description</h2></div><div><p><strong>At Bosch, we shape the future by inventing high-quality technologies and services that spark enthusiasm and enrich people’s lives. Our promise to our associates is rock-solid: we grow together, we enjoy our work, and we inspire each other. </strong><strong>Join in and feel the difference.</strong></p><p>The Robert Bosch GmbH<em> </em>is looking forward to your application!</p></div></section><section><div><p></p><h2>Job Description</h2></div><div><p>Detecting objects in the surroundings is one of the first essential tasks for autonomous driving. For object detection, radar is one of the most used perception sensors. By using radar, the environment can be represented as a point cloud. For extracting objects from this radar point cloud, two kinds of methods are currently studied: classical methods associate the points based on their features, predict possible objects using physical models, and track them using Kalman filters; Deep-Neural-Network (DNN)-based models recognize patterns in the point cloud and extract possible objects as bounding boxes after proper training. Both methods have their strengths and limitations depending on the use cases. Therefore, an optimized combination of these two methods is essential for improving detection accuracy.</p><ul><li>In this Master thesis, you will first evaluate the classical and deep-learning-based methods for radar object detection based on different scenarios. Afterward, you will implement a method – so called hybrid modelling – to combine these two methods based on their current detection quality and integrate the two methods with further multi-object tracking. The implementations will be evaluated with datasets recorded in real-world scenarios. Contribution to academic publications is possible and welcomed.</li><li>You study and analyze the state-of-the-art algorithms for object detection, DNN uncertainty, and data association and you evaluate the classical and DNN object detection methods using radar point cloud data.</li><li>You exploit the advantages of these two different approaches by developing novel algorithms to combine these models.</li><li>Furthermore, you can evaluate your implementation using advanced sensor setups and challenging real-world datasets.</li><li>Beyond that, you can support the integration of your algorithms in the ROS workspace of the test vehicle.</li><li>You work in an agile and diverse research team and collaborate closely with experts in perception and deep learning. Last but not least, you benefit from the powerful resources of Bosch to accelerate your research.</li></ul></div></section><section><div><p></p><h2>Qualifications</h2></div><div><ul><li><strong>Education: </strong>Master studies in the field of Computer Science, Electrical Engineering, Mechanical Engineering, or comparable</li><li><strong>Experience and Knowledge: </strong>good programming skills in Python or C++, prior experience with Deep Learning frameworks such as TensorFlow or PyTorch, perception, or ROS is a plus</li><li><strong>Personality and Working Practice:</strong> independent, structured and communicative team player</li><li><strong>Enthusiasm: </strong>passionate about innovation and technologies, motivated to continuously expand skills and knowledge in this area</li><li><strong>Languages:</strong> very good English</li></ul></div></section><section><div><p></p><h2>Additional Information</h2></div><div><p><strong>Start:</strong> according to prior agreement<br>
<strong>Duration:</strong> 6 months</p><p>Requirement for this thesis is the enrollment at university. Please attach a motivation letter, your CV, transcript of records, examination regulations and if indicated a valid work and residence permit.</p><p>Diversity and inclusion are not just trends for us but are firmly anchored in our corporate culture. Therefore, we welcome all applications, regardless of gender, age, disability, religion, ethnic origin or sexual identity.</p><p><strong>Need further information about the job?</strong><br>
Jiaying Lin (Business Department)<br>
+49 173 3263759</p></div></section></div>

Master thesis Hybrid Modeling for Object Detection using Radar Point Clouds in Autonomous Driving

remote

about 1 month ago