CMSA Computer Science for Mathematicians: Generation by Decomposition


View Calendar
October 6, 2020 11:30 am - 12:30 pm
via Zoom Video Conferencing

Hadar Averbuch-Elor - Cornell Tech

Deep learning has revolutionized our ability to generate novel images and 3D shapes. Typically neural networks are trained to map a high-dimensional latent code to full realistic samples. In this talk, I will present two recent works focusing on generation of handwritten text and 3D shapes. In these works, we take a different approach and generate image and shape samples using a more granular part-based decomposition, demonstrating that the whole is not necessarily “greater than the sum of its parts”. I will also discuss how our generation by decomposition approach allows for a semantic manipulation of 3D shapes and improved handwritten text recognition performance.