On October 7,  Dr. Stefano Cagnoni from the Department of Computer Engineering of the University of Parma will give a talk.

Date: 7th Oct. 2009, 10-11:00 am

Place: Room 174, Snelius Building, Niels Bohrweg 1, Leiden (the Computer Science department of Leiden University)

Title: “GPU-based implementation of Particle Swarm Optimization algorithms”

Abstract: Particle Swarm Intelligence is a modern artificial intelligence discipline that is concerned with the design of multi-agent systems with applications. Instead of a sophisticated controller that governs the global behavior of the system, the swarm intelligence principle is based on many unsophisticated entities that cooperate in order to exhibit a desired behavior, for example, the foraging behavior that ants exhibit when searching for food. While ants employ an indirect communication strategy via chemical pheromone trails in order to find shorted paths between their nest and food sources.

Scientists have applied these principles to new approaches, for example, in optimization and the control of robots. In this talk, Dr. Stefano Cagnoni will describe and compare possible approaches to GPU-based implementation of Particle Swarm Optimization (PSO) within the NVIDIA CUDA environment. In particular, the talk will highlight the main problems that must be tackled to reach the highest possible degree of parallelization and discuss the pros and cons of different approaches that can be followed to solve them. Results achieved on a sample real-time object detection toy problem will also be described.

Short CV: Dr. Stefano Cagnoni graduated in Electronic Engineering at the University of Florence in 1988 where he has been a PhD student and a post-doc until 1997, working in the Bioengineering Lab of the Department of Electronic Engineering. He received the PhD degree in Bioengineering in 1993. In 1994 he was a visiting scientist at the Whitaker College Biomedical Imaging and Computation Laboratory at the Massachusetts Institute of Technology (MIT).  Since 1997, he has been with the Department of Computer Engineering of the University of Parma, where he is currently Associate Professor. Dr. Stefano Cagnoni’s main research interests are: Computer vision, Robotics, Evolutionary Computation and Neural Networks.