Speech and natural language processing is a subfield of artificial intelligence used in an increasing number of applications; yet, while some aspects are on par with human performances, others are lagging behind. This course will present the full stack of speech and language technology, from automatic speech recognition to parsing and semantic processing. The course will present, at each level, the key principles, algorithms and mathematical principles behind the state of the art, and confront them with what is know about human speech and language processing. Students will acquire detailed knowledge of the scientific issues and computational techniques in automatic speech and language processing and will have hands on experience in implementing and evaluating the important algorithms.
Topics:
- speech features & signal processing
- hidden markov & finite state modeling
- probabilistic parsing
- continuous embeddings
- deep learning for language-related tasks (DNNs, RNNs)
- linguistics and psycholinguistics
- comparing human and machine performance
Basic linear algebra, calculus, probability theory
The courses take place on monday, from 9am to 12am. Be on time!
- #1 Jan 22 (235B). Introduction (Sagot & Dupoux)
- #2 Jan 29 (Dussane). ASR1: Features and Acoustic Models (Dupoux & Zeghidour) + TD
- #3 Feb 05 (Dussane). ASR2: Language Models (Dupoux, Zeghidour, Riad) + TD
- #4 Feb 12 (Toqueville). NLP1: Language processing in the wild (Sagot) + TD
- #5 Feb 19 (Dussane). NLP2: Formal languages (Sagot) + TD
- #6 Feb 26 (Dussane). NLP3: Parsing (Sagot) + TD
- #7 Mar 05 (Dussane). Translation (Guest: Schwenk)
- #8 Mar 12 (Actes). Perspectives (Sagot & Dupoux)
(235B): 29 rue d'Ulm, Paris 75005, 2nd Floor, left. (Dussane): 45 rue d'Ulm, Paris 75005, Amphi Dussane, Ground Floor, left. (Actes): 45 rue d'Ulm, Paris 75005, Salle des actes, 1st Floor, right. (Toqueville): 61 Avenue du Président Wilson, 94230 Cachan, Batiment d'Alembert, Amphi Tocqueville.
The course materials (PDFs, etc.) are listed in the subdirectories numbered #1 .. #8.
The practical assignments are given on the courses #2 to #6. For the assignments, students are provided with the necessary data and Python code in the form of a disk image (.iso) to be mounted and booted from a virtual machine. They will hand in their source code and a max two page report, detailing their work, the difficulties encountered and the results.
Please install VirtualBox (https://www.virtualbox.org/) on your machine and load up the provided disk_image.iso before coming to the first TD.
During the courses, we will use on-line quizzes (on the smartphone/computer) to probe comprehension and trigger discussion.