Francesco Donnarumma
DBLP search
ResearchGate Profile
Google Scholar Profile
Microsoft Academic Research Profile Profile
Linkedin Profile
Linkedin Profile
Facebook Profile
Twitter Profile
MySpace Profile


I have been involved in several projects: International (HSFP - Beyond simple choices: computational and neuronal mechanisms for complex spatial behaviors), European (ETHICBOTS, DEXMART, HUMANOBS, Goal-Leaders), National (Cyber Brain project), Regional (MONDIEVOB, EVODIALIFT, CAWSYS).

My expertise spans:

*) Computational Modelling of brain functions by Dynamic Neural Networks. One of my principal research product is related to a programmable fixed-weight dynamic neural network architecture (see “Programming in the brain: a Neural Network theoretical framework“) built on a biologically plausible recurrent Artificial Neural Networks (ANNs) model - Continuous Time Recurrent Neural Networks (CTRNNs) capable of exhibiting multiple behaviours and “fast“ switching among them. A number of works on this model were developed (see ,“How and over what timescales does neural reuse actually occur?”, "A Programmer - Interpreter Neural Network Architecture for Prefrontal Cognitive Control", “A Robotic scenario for programmable fixed-weight Neural Networks exhibiting multiple behaviors” and "Learning programs is better than learning dynamics: a Programmable Neural Network Hierarchical Architecture in a multi-task Scenario").

*) Cognitive Robotics and Neural Networks for Robotic Systems. My research focused on computational models that includes biological aspects of the circuit including the hippocampus, the ventral striatum and the sensory-motor cortex and explains mechanistically how it may be used to imagine and evaluate future events (see: “Mental imagery in the navigation domain: a computational model of sensory-motor simulation mechanisms”), models for biologically (rat - monkeys) inspired strategic planning by probabilistic inference (“Divide et Impera: subgoaling reduces the complexity of probabilistic inference and problem solving”) and probabilistic Bayesian models for social interaction between agents (“Human Sensorimotor Communication: A Theory of Signaling in Online Social Interactions” and "Interactional Leader-Follower sensorimotor communication strategies during repetitive joint actions").

*) Machine Learning, Neural Network theory and Evolutionary Computing. In “Model Predictive Control Strategy Based on Differential Discrete Particle Swarm Optimization” techniques based on the discretization of Particle swarm optimization were developed. Differential Evolution strategies were introduced in the CTRNN framework in “CTRNN Parameter Learning using Differential Evolution”. Evolutionary techniques are also used in a Robotic scenario in "Learning programs is better than learning dynamics: a Programmable Neural Network Hierarchical Architecture in a multi-task Scenario". A neural network architecture was proposed in “An action tuned Neural Network Architecture for hand pose estimation” for vision based hand pose estimation. Moreover, a novel method was developed in order to reconstruct 3D environments for Unmanned Aerial Vehicles (UAVs) (see “Fast incremental clustering and representation of a 3D point cloud sequence with planar regions”). In "Dynamic Network Functional Comparison via Approximate-bisimulation", a new method for the comparison of dynamic networks is presented.

I got several didactic roles in courses at Informatics Department of University Federico II of Naples in the period 2007-2012 (e.g. Laboratory of Programming, Minds and Machines, Machine Learning, Neural Networks, Computer Vision). In 2014, I was header of the Course in "Analysis and interpretation of brain signals for Brain-Computer Interfaces applications" for Italian National Project 'Cyber-Brain project' aiming at forming specialized researchers for the constitution of a BCI centre of excellence in the macro-region south of Italy.