Talks

Prof. Yossi Keshet

Dr. Yossi Keshet

From Classical Speech Processing to Deep Learning

In this talk, I will present a series of works aimed at replacing the traditional stochastic signal processing algorithms with a deep learning approach. First, I will discuss the difficulties in estimating pitch (fundamental frequency) and formant frequencies using deep neural networks, and will present our successful attempts to tackle these tasks. Then, I will present an algorithm for unsupervised time-scale modification of speech. I will conclude the talk with a novel method to Speech Steganography, namely hiding a secret spoken message within an ordinary public spoken message.

The talk is dedicated to our friend Yoav Medan who passed away last month and whose signal processing algorithmic thinking inspired my work in the field.

Link to personal website

Elias Nehme

Elias Nehme

Towards Intelligent Microscopes with Deep Learned Optics

In biological imaging, fast acquisition of depth information is crucial e.g. for accurate 3D tracking of sub-cellular elements and for 3D super-resolution. In this talk, I will present a series of works enhancing the success of snapshot depth sensing in the revolutionary field of single-molecule localization microscopy (Nobel Prize in Chemistry 2014). Specifically, I will present an approach for jointly designing the “optics” of the microscope and the 3D reconstruction algorithm, by using deep learning. Throughout the talk I will demonstrate our approach experimentally with super-resolution reconstructions of mitochondria and volumetric imaging and tracking of fluorescently labeled telomeres in live cancer cells

Link to personal webpage