Cairo AI Meetup #4 - AI in Cars

Hello again AI Geeks..

It's been 9 long months since our last meetup, but...Affectiva is back with the 4th edition of Cairo AI Meetup on Sept 22nd with collaboration and support from Dell EMC (host and sponsor) and Alama (co-organizer).

This meetup will feature speakers from the industry working towards the Cars of the future! Join us on September 22nd to make new connections as well as listen to enlightening talks from Affectiva and Valeo.

We will have longer networking breaks, with semi-moderated discussions to help you make the most out of this community!

Places are limited, for registration, you must RSVP and secure a spot + Fill the registration form : http://bit.ly/cairoAImeetup4 (both are required)

Agenda :
01:00 PM - Registration at Venue (you Must bring your national ID / passport)
01:30 PM - Welcome note
01:45 PM - Speech emotion recognition inside the Car - Youssef Emad (Affectiva)
02:30 PM - Networking break + Group discussions
03:15 PM - YOLO3D: End-to-end 3D Object Detection from LiDAR Point Cloud - Sherif Abdelkarim (Valeo)
04:00 PM - Networking break + Snack
05:00 PM - An automatic image captioning model - Ramy Wafa (Affectiva)
5:30 PM - Closing

Abstracts

Speech emotion recognition inside the Car - Youssef Emad (Affectiva) :
Emotion Recognition from speech can be a very challenging task especially in-car. We will discuss Affectiva's classification pipeline to overcome various challenges including the lack of any real or academic in-car emotion datasets. We will also explore multiple modeling approaches from CNNs and RNNs to Transfer Learning and Multiple Instance Learning.

YOLO3D: End-to-end 3D Object Detection from LiDAR Point Cloud - Sherif Abdelkarim (Valeo) :
This talk starts by giving a very brief intro to object detection and will differentiate between single shot and two stage detectors. Then the talk will go on to explaining the YOLO architectures and lastly will go into the changes we made to the architecture to tailor it to our specific problems at Valeo. Our paper detailing the changes we made to the architecture along with the results are accepted in ECCV workshop 2018.

An automatic image captioning model - Ramy Wafa (Affectiva) :
We will talk about a model combining CNNs and RNNs to produce an automatic image captioning model. We use CNNs to distill image features and feed that into an RNN to produce a description of that image.

See you soon :)