This course introduces the fundamentals of speech and speech technologies, their real-world applications, and provides hands-on experience with relevant speech processing toolkits. This course is also offered as an elective for PG-diploma students.

Lectures-Tutorials-Practicals-Credits: 4-0-0-4

Week 1: Introduction to speech and signal processing; fundamentals of sampling, Fourier analysis, and speech production and perception.

Week 2: Speech analysis techniques; perceptual features, filtering, and feature extraction including MFCCs.

Week 3: Statistical modeling for speech; Gaussian models, classification, and mixture models.

Week 4: Vector quantization and sequence modeling; Markov chains and the Viterbi algorithm.

Week 5: Hidden Markov Models (HMMs); sequence inference and probabilistic modeling.

Week 6: Speech recognition systems; feature extraction and HMM-GMM based approaches.

Week 7: Language modeling and introduction to neural networks; n-grams and deep learning basics.

Week 8: Feedforward neural networks and training; backpropagation and introduction to CTC.

Week 9: Recurrent neural networks and sequence modeling; applications and limitations.

Week 10: Advanced neural architectures; word embeddings, sequence-to-sequence models, and convolutional neural networks.

Week 11: Transformers and modern deep learning approaches; self-attention and self-supervised learning.

Week 12: Speech synthesis and speaker recognition; end-to-end systems, verification, and diarization.