Neuromorphic computing for ai solutions and neuro-robotics

The course in Neuromorphic computing will provide the student with knowledge in the field of computational neuroscience and the state-of-the-art understanding of biological sensorimotor systems, allowing their implementation in artificial computational intelligence systems, and on parallel computing systems for faster and biomimetic processing of sensory information.
The first part of the course focuses on the functioning of human sensorimotor systems. Beginning from the working principles of sensory transduction, through the central processing and integration of these signals, to the production of motor responses and their adaptive capabilities. Neurophysiological understanding of these functions is translated into neuromorphic mathematical models capturing and reproducing the processing and the role of specific anatomical structures involved in a sensorimotor system. Examples relative to specific brain structures will be presented and tools to implement, simulate and interpretate these models in both physiological and pathological states will be provided. The exam will be held orally.
The second part of the course focuses on Parallel Programming and aims at providing the students with the fundamentals of programming systems with manycore and multicore processors and the implementation of neuromorphic models particularly making use of Artificial Intelligence techniques.
Specifically, the course includes a portion focused on GPU programming leveraging the CUDA development environment.
The exam is designed in the form of a project to be developed individually.
A list of possible projects (by complexity and development time) will be proposed by the lecturers and each student can choose from them.

Knowledge of the concept of biological control systems governing our physiological behavior and how to represent and implement them in an artificial computational intelligence system.
Completion of the parallel programming project will result in the following learning outcomes:
- preparation of a concise report on the problem addressed and the implementation strategies adopted
- writing clear and well-documented code
- design of a test phase in which the correctness of the code is demonstrated on significant benchmark instances
- analysis of the performance (speedup and profiling in general) achieved by the parallel versus sequential algorithm.