Learning Deeply

Speaker:  Jon G Peddie – Tiburon, CA, United States
Topic(s):  Architecture, Embedded Systems and Electronics, Robotics

Abstract

Can an AI make a movie?
 
The first step was training it on trailers of 100 horror movies. Each trailer was segmented into “moments”
 
Each scene was tagged with an emotion from 24 emotions and labels from 22,000 scene categories, such as eerie, frightening and loving;
 
An audio analysis of the ambient sounds to understand the sentiments of those scenes;
 
An analysis of each scene’s composition that make up suspense/horror movie trailers.
 
The analysis was performed on each area separately and in combination with each other using statistical approaches. The system now “understands” the types of scenes that categorically fit into the structure of a suspense/horror movie trailer.
 
AI has been there for many years,  and it can be applied in many places. 
 
GPUs have enabled DL, and  therefore, AI
 
Data scientists in industry and academia have been using GPUs for machine learning to make groundbreaking improvements across a variety of applications including:
 
Deep Learning has enabled many practical applications of Machine Learning and by extension the field of AI. 
 
Deep Learning breaks down tasks in ways that makes all kinds of machine assists seem possible - Driverless cars, better preventive healthcare, even better movie recommendations
 
AI is the present and the future. 
 
With Deep Learning’s help, AI may even get to that science fiction state we’ve so long imagined
 
It’s not just about finding cats


About this Lecture

Number of Slides:  31
Duration:  40 minutes
Languages Available:  English
Last Updated: 

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.