Natural Language Processing & Speech Recognition

Course details

Natural Language Processing (NLP) and Speech Recognition are at the forefront of modern AI applications, from chatbots to voice assistants. This course will provide students with a comprehensive understanding of the foundational and advanced concepts of NLP and Speech Recognition. Through a mix of theoretical concepts and hands-on labs, students will be prepared to develop and optimize real-world applications.

Start Date
TBD
Duration
14-15 weeks
Prerequisites
Foundational knowledge in AI & ML, and basic programming skills.
Learning Objectives
  • Grasp the fundamental concepts of NLP and its application in AI.
  • Understand and implement Speech Recognition techniques.
  • Engage with text and speech data, preprocessing, and analysis.
  • Create models for sentiment analysis, entity recognition, and machine translation.
  • Apply advanced techniques in deep learning to NLP & Speech Recognition tasks.
Key Features
  • Practical labs using popular NLP libraries such as NLTK, spaCy, and HuggingFace.
  • Analysis of real-world data sets.
  • Exposure to the latest in voice technology applications.
  • Exploration of ethical considerations in NLP & Speech Recognition.
Learning path
  • Unit 1: Introduction to NLP (3)
  • Unit 2: Classic NLP Techniques (5)
  • Unit 3: Advanced NLP with Deep Learning (4)
  • Unit 4: Introduction to Speech Recognition (3)
  • Unit 5: Deep Learning in Speech Recognition (4)
  • Unit 6: Applications & Use Cases (4)
  • Unit 7: Challenges & Future Directions (4)
  • Unit 8: Wrap-up and Review (1)

Topics covered

  • Overview of NLP
  • Language Models: n-grams, Bag of Words
  • Text Preprocessing and Normalization

  • Part-of-Speech Tagging
  • Parsing and Syntax Trees
  • Named Entity Recognition
  • Sentiment Analysis
  • Topic Modeling with LDA

  • Word Embeddings: Word2Vec, GloVe
  • RNNs for Text
  • Transformers & BERT
  • Machine Translation

  • Fundamentals of Speech Processing
  • Feature Extraction in Speech: MFCCs
  • Hidden Markov Models (HMMs)

  • CNNs for Speech Data
  • DeepSpeech & RNN Transducer Models
  • Attention Mechanisms in Speech Recognition
  • State-of-the-Art Models and Benchmarks

  • Chatbots and Virtual Assistants
  • Voice Assistants: Siri, Alexa, Google Assistant
  • Audio-to-Text Services
  • Multimodal Models (Text, Image, Speech)

  • Handling Multilingual and Diverse Dialects
  • Bias and Fairness in NLP
  • Ethical Considerations in Speech Recognition
  • The Future of Interactive AI

  • Course Review and Exploration of Current Research

Grading:

  • Class Participation: 10%
  • Weekly Assignments: 30%
  • Group Projects: 20%
  • Mid-Term Examination: 20%
  • Final Examination: 20%

Reading Material: Combination of standard textbooks, recent research papers, and articles from the domain of NLP & Speech Recognition.

Assignments: Hands-on labs and projects, focusing on real-world problems and data sets.

Faculty: Expert in the fields of NLP and Speech Recognition.

Teaching Instructors: TBD