BYU

Abstract by Seong-Eun Cho

Personal Infomation


Presenter's Name

Seong-Eun Cho

Degree Level

Undergraduate

Abstract Infomation


Department

Mathematics

Faculty Advisor

Tyler Jarvis

Title

Image encoding of English sentences

Abstract

One challenge in encoding natural language sentences is their variable length. In my research, I attempt to encode sentences into a fixed-length vector using Transformer layer and image encoding with the power of neural nets. The encoding is done in an auto-encoding procedure, and this may open new ways to encode meaning in sentences and be able to sample new sentences from a vector space.