Accelerators for object detection and tracking
At the IIS we are working on an ultra low-power multi-processor shared-memory cluster realized in 28nm SOI technology. One of the applications for this system is in real-time video applications, expecially in systems where objects and/or people need to be identified and tracked in a live video stream.
In this project, you will design a hardware accelerator for image processing applications that can be included in the multi-processor system, improving the energy efficiency of the system. The accelerator will be designed so that it can connect to the existing infrastructure of the current system.
By doing this project, you will gain experience in designing a complex heterogeneous computing system in nano-meter technology. You will learn recipes to create a balanced mix of configurable special function accelerators and parallel programmable platforms to achieve maximum energy efficiency. You will also become part of the team that is currently developing the multi-processor shared-memory cluster.
- Looking for 1-2 Semester/Master students
- Contact: Frank K. Gurkaynak
- VLSI I
- VLSI II (recommended)
- 20% Theory
- 40% ASIC Design
- 40% EDA tools