Beirut AI Quarterly Meetup #4
Saturday, December 1st at Beirut Digital District marked the 4th and last quarterly Beirut AI Meetup for 2018. For those of you not familiar with the events, Beirut AI hosts quarterly meet-ups in order to showcase local AI applications in Lebanon. Every event, we select a few speakers and AI practitioners to share with the audience their experiences applying AI. They recap their journey on how they applied it in their work, the successes and failures they faced along the way, and the lessons and knowledge they gained in doing so.
Taking the stage first, Rami Santina president and founding member of the Syndicate of Computer Sciences in Lebanon (SCSL) discussed SCSL and how to apply to become a member. We are partnering with SCSL and will keep you posted on future event collaborations.
The speaker lineup for the 4th Quarterly Meet-up was as follows:
Ramzi Karam, Senior Machine Learning & Backend Engineer, Anghami
In “From billions of streams to better recommendations,” Ramzi gave the audience insight into Anghami’s recommendations infrastructure. Using data that surfaces from users listening to 100 years worth of music, as well as deep learning, music information retrieval, matrix factorization and Word2vec algorithms, Anghami’s recommendations transcend the ordinary ‘top of the charts’ playlists and give users tunes that they’re sure to enjoy.
Dr Joe Tekli, Assistant Professor, Department of Electrical & Computer Engineering, LAU
MUSEC, is a framework that composes music, developed by Dr. Tekli’s graduate student, Ralph Abboud. In ‘MUSEC, a framework for Music Sentiment-based Expression and Composition,’ Tekli explains how they experimented with the limits of the human brain and AI.
The way MUSEC works is it assess the emotional percentage of a piece. For example, it concludes that a musical piece is composed of 30% happiness, 5% sadness, etc. Users can input chosen emotions, and it composes music based on those emotions. The framework then creates an audio file made up of these notes that sound almost too real to be true. Dr Tekli gave a brief demo of the result that stunned the audience thanks to its musical perfection.
Abdelrahman Ghallab, Data Scientist, Numbase Group
In Deep Learning in Medicine, Ghallab gave an interesting talk about his groundbreaking projects using computer vision and AI to detect early stages of Melanoma, the most frequent type of skin cancer. Using a more affordable and user-friendly method that can be accessible through a smartphone, they allow people to scan their skin and get it sent to the cloud, in order to determine if it really is cancer.
Beirut Digital District was buzzing with talks of AI that weekend, as Techstars Global Startup Weekend AI (GSWAI) also took place at the same time. As co-organizers, Beirut AI gave a brief introduction about Techstars global network’s startup weekend, the agenda for the weekend, and why the global AI edition was important.
You can catch the 4th quarterly meet-up in full below.
And don’t forget to follow us to keep an eye out for future events and meet-ups!
Facebook: https://www.facebook.com/BeirutCityAI/
Instagram: https://www.instagram.com/beirut.ai/
