Objectives and outcomes
Acquiring the knowledge necessary to apply artificial intelligence and machine learning algorithms to the
analysis and processing of image signals, time series, video and audio, as well as biomedical signals.
The ability to analyse large data obtained from the Internet of Things. Students are trained to analyse,
implement, apply and evaluate artificial intelligence algorithms in image signal processing, time series,
video and audio signals, as well as to analyse biomedical signals. They know how to apply advanced
artificial intelligence techniques to mass distributed systems, especially the Internet of Things.
Lectures
Artificial neural networks with direct signal propagation, recurrent neural networks, convolutional neural
networks. Image, video and audio signal processing using neural networks. Handwriting and speech
recognition. Face recognition. Video segmentation and boundary frame detection. The real-time position of the hand. Convergence of augmented reality, intelligent virtual agents and the Internet of Things. A hybrid approach to image segmentation in the Internet of Things. Big data analytics for the Internet of Things. Time series analysis and prediction. Machine learning techniques in the
analysis of biomedical signals (ECG, EEG, EMG). Noise removal, property extraction, dimensionality
reduction (PCA, ICA, KPCA, MSPCA), entropy and other statistical measures.
Practical classes
Implementation of neural networks with direct signal propagation, recurrent neural network, convolutional
neural networks. Image, video and audio signal processing using neural networks. Time series analysis
and prediction. Using the TensorFlow and Keras tools. Extraction of boundary frames using Torchvision
and Torch tools. Implementation of supervised and unsupervised learning in the Internet of Things.
Distributed processing of the Internet of Things via Apache Spark using MLlib and H2O.ai platforms.
Case studies for Personal IoT, Industrial IoT and smart cities.