Big data for remote sensing applications

Mesay Belete Bejiga

Big data, Convolutional neural networks, Remote sensing, Unmanned aerial vehicles

The availability of large remote sensing data from satellites and airborne sensors have brought challenges and opportunities in remote sensing applications. Those data are being used for several applications, such as urban area monitoring, disaster management, and assisting search and rescue operations. However, the massive amount of data available requires efficient methods that can process and analyze the data at a reasonable time. In this research work, we will develop novel image processing methods that can analyze and extract information from massive remote sensing image datasets for use in disaster management and search and rescue applications.




Information Retrieval from Neurophysiological Signals

Pouya Ghaemmaghami Tabrizi

Pattern Recognition, Machine Learning, Signal Processing, Computer Vision

Brain Computer Interface (BCI) has recently become a hot topic outside neuroscience and rehabilitation engineering communities. Other disciplines such as computer science have already started to contribute to the field by applying signal processing approaches to the brain data. This research aims at decoding brain signals in order to retrieve multimedia contents by using novel machine learning algorithms.




User-centric affective multimeda tagging

Mojtaba Khomami Abadi

emotion, personality, neuro-physiology signals

At the University of Trento, Mojtaba's first research activities were on brain-decoding of subjects in response to affective multimedia content using pattern analysis on neuro-physiological signals. Mojtaba had great access to cutting edge brain imaging technologies including the MEG brain sensor at the Center for Mind/Brain Sciences (CIMeC), Trento. He had major contributions in the development of the very first publicly available MEG-brain signals dataset for decoding human emotions in response to emotional Hollywood videos and music videos. He developed algorithms that effectively decoded human emotions from MEG brain signals being published in the IEEE Transactions on Affective Computing. Mojtaba also committed to the development of some more publicly available datasets for recognizing human emotions and personality from commercial portable sensors where he also developed AI algorithms for emotion and personality recognition using commercial sensors.