Privacy-Preserving Encrypted Deep Learning Presentation

Fully Homomorphic Encryption (FHE) is the holy grail of encryption. Data it encrypts is manipulatable/ computable while it is still encrypted as a cyphertex. This quantum resistant encryption creates a way for third parties to interact with highly sensitive data without any possibility of data leaks, since they never require the ability to decrypt the cyphertext, but can still use it to spit out encrypted answers to some of the most difficult and pressing problems; Diagnosis, forecasting, automation etc. This presentation will introduce FHE applied to deep learning, show you in brief how it works, and point you in the right direction if you should want to learn more.

Questions

To ask a question please create a new issue using this link: https://github.com/DreamingRaven/encrypted-deep-learning-presentation/issues/new

to build the presentation

install docker and docker-compose: https://docs.docker.com/engine/install/ https://docs.docker.com/compose/install/

then while in this directory run:

docker-compose up

The presentation pdf will be built inside the newly created "built" directory.

Previously presented

This presentation has been used in some form to present at:

Internet of Food Things Virtual Workshop: How Technology Can Facilitate Data Sharing In The Agri-Food Sector (2021-03-01)
University of Lincoln School of Computer Science: Research Seminars (2021-02-10)
New Scientist Live 2020 (2020-11-28)