Machine Learning for Speech and Audio Processing
Lecturer: Prof. Dr.-Ing. Peter Jax
Contact: Lars Thieling, Erik Fleischhauer
Type: Master lecture
Credits: 4
Registration via RWTHonline
Course language: English
Material:
Lecture slides and Exercise problems will be published in RWTHmoodle.
The lecture "Machine Learning for Speech and Audio Processing (MLSAP)" addresses especially students of the Master's program "Electrical Engineering, Information Technology and Computer Engineering". The formal connection to the module catalogs can be found at RWTHonline.
Content
In this one term lecture the fundamental methods of machine learning with applications to problems in speech and audio signal processing are presented:
- Fundamentals of Classification and Estimation
- Bayesian Probability Theory: Classification and Estimation
- Feature Extraction Techniques
- Modeling of Statistical Distributions
- Basic Classification Schemes
- Probabilistic Models
- K-Means Clustering
- Gaussian Mixture Models (GMMs)
- Expectation-Maximization (EM) Algorithm
- Modeling Sequential Data
- Hidden Markov Models (HMMs)
- Estimation and Classification with HMMs
- Linear Dynamical Systems (LDS)
- Non-Negative Matrix Factorization (NMF)
- Neural Networks and Deep Learning
- Elements of Neural Networks
- Feed-Forward Neural Networks
- Training of Synaptic Weights: Backpropagation and Stochastic Gradient Descent (SGD)
- Specialized Network Architectures: CNNs, RNNs, LSTMs
- Advanced Learning Techniques
Exercises are offered to gain a deeper understanding on the basis of practical examples.