Abstract

Deep Learing for Image Understanding

In this session, Amir will describe deep learning models that can use context to make predictions. Specifically, he will describe recent models for image annotation and text categorization. The models we consider iteratively refine their prediction by considering how each prediction components is influenced by other components...


Session No: SIL8135
Speaker: Amir Globerson
Type:

Deep Learning and AI

Date: Thursday - October 18, 2018 02:00 PM - 02:45 PM
Location: Hall K/L
Topics: Deep Learning and AI, Autonomous Vehicle Technologies, Accelerated Data Science, Professional Visualization
Industry: General

In this session, Amir will describe deep learning models that can use context to make predictions. Specifically, he will describe recent models for image annotation and text categorization. The models we consider iteratively refine their prediction by considering how each prediction components is influenced by other components. For example, in image annotation, there exists strong correlations between the identity of the objects in the image, and we show how our model can capture these effects. Our model produces state of the art results on the challenging problem of annotating images with objects and relations.