Student Assistant (Hiwi) for the development of traffic light detection for autonomous vehicles with neural networks
The goal of your task is to develop a system that can recognize camera-based traffic light signals. Neural networks such as Vision Transformer or modern CNNs should be used for this purpose.
Since the network is also intended to be used in our autonomous vehicles in the long term, speed/efficiency plays a crucial role. Your tasks:
- Research the current state of the art for traffic light detection and object detection in general.
- Conceptualize, create, and train the new network on the research group's hardware.
- If necessary, record new data with our autonomous vehicles. The test field Baden Württemberg in Karlsruhe allows direct communication with traffic lights, so time-consuming labeling tasks is likely not necessary.
- Integrate the network into the perception component of the FZI's driving stack for autonomous vehicles.
- Test and evaluate the network in the traffic of Karlsruhe and on campus east.
You should have:
- Experience in the field of machine learning, particularly deep learning for computer vision.
- Excellent knowledge of Python. Knowledge of C++ can be advantageous.
- Experience in PyTorch or good knowledge of other equivalent frameworks (Tensorflow, Jax).
- Knowledge of ROS is beneficial but not mandatory.
- Interest and commitment to the subject matter.
These tasks can also be carried out as part of a master's thesis.
If you are interested, please email me with your current transcript and a brief description of why you would like to take on the position.
Nikolai Polley∂kit edu
Student Assistants / Tutors
No information available