Artificial Intelligence in Image and Signal Analysis – AI-ISA

Research Activity

The AI-ISA Group has a wide experience in the field of Computer Vision, Pattern Recognition and Signal Analysis, as well as skills in the field of Machine Learning.

The group’s research activities focus on:

– Development of methods and design of new networks and learning algorithms for feature extraction/identification/classification of patterns in images and signals.

– Development of models for the interpretability of machine learning systems, to understand/explain the obtained prediction or classification.

– Development of models based on end-to-end Machine Learning systems in the context of Generative Adversarial Networks for the generation of new patterns with characteristics similar to those of the real data in specific application domain.

– Development of Machine Learning models for mobile devices designed to provide low latency inference performance at an extremely low cost in terms of processing capacity, memory and energy consumption.

Goals

The AI-ISA Group has three main goals:

– Design and development of Artificial Intelligence (AI) models and methods for the automatic analysis of images and signals which due to their nature are difficult to inspect by a human operator or which require highly specific skills in order to extract the semantic content.

– Study of methods for the interpretation and explanation of the decision-making mechanisms underlying complex AI models.

– Design of AI models for image and signal analysis in environments with limited resources.

Application Fields

Particular interest for the application fields of Medical/Biological Imaging and Healthcare.

• MOLIM ONCOBRAIN LAB – Innovative methods of molecular imaging for oncological and neurodegenerative diseases
• Development of an integrated radiomic and phenotypic system for the diagnosis, prognosis and personalized therapy of head and neck cancer therapy. Technological platform: eMORFORAD-Campania