Welcome to EE545 Embedded Deep Learning
The course will cover various correlated concepts starting with an overview of deep learning architectures and following with how to embedd deep learning systems on harwdare, their challenges and the state of the art methods. The course will focus mainly on understanding how to to run and optimize neural architecures on actual harware.
Due to the extraordinary situation we are all in, and with the permission of the Çankaya University Senate, we are changing the grading of the course for this semester.
- Homework (1): %10
- Midterm (none):
- Final (1): %40
- Project (1): %50 (%25 online Presentation + %25 Project Report)
- either a rigourous theoretical analysis: Matlab or Python (using a deep learning library of your choice)
- or a practical Analysis (or sometimes both)
- FPGA Implementation (e.g. PYNQ-Z2)
- GPU Implementation (e.g. Jetson NVIDIA)
- CPU implementation (Arm etc.)
Homework: We already had it.
Final Exam: will be held some time in June. The Finals Calendar will be announced officially as usual. The departmental board agreed on (but not strictly enforced) the idea of a "classical exam on Zoom with cameras on", so that I may be able to answer you questions during the exam.
Final Exam: will be held some time in June. The Finals Calendar will be announced officially as usual. The departmental board agreed on (but not strictly enforced) the idea of a "classical exam on Zoom with cameras on", so that I may be able to answer you questions during the exam.
Project Presentations: will be held on a Zoom session, to which everybody will strictly attend. Each one of you will make a Zoom presentation of 10 minutes (~10 slides) by sharing your screen, by which you will explain your project. This will be sometime in the first week of June.
Project Reports: will be emailed after the presentation, up until June 15. I need all analyses, experiments (if any) clearly reported with a conclusion of your studies.