Impactful AI - 3rd Meetup

In this meetup series, we want to inspire everyone to use or implement Machine Learning and AI not only to serve great causes but also to have a demonstrable positive impact on society. We’ll be talking about how we can use AI to do just that, through applying it to health, education, democracy, nutrition and more, but also how we can technically implement that while ensuring what we build is ethical.

Our speakers will be:

Kerem Sozugecer, CTO @ DeepZen
> Text to Speech for Social Impact

Abstract: DeepZen is a Text to Speech company specialised in creating ultra-realistic emotional and expressive human voice.
We create audiobooks and audio content using AI.
A typical human narrated book get created in 6-8 weeks where we can do it within days.
DeepZen also utilizes its technology for social impact:
We have partnered up with IBM to do a pilot in one of the UK schools. The goal is to use DeepZen's technology to increase the number of audio content for children with dyslexia and visually impaired.
We are also doing a pilot with Welsh NHS to convert leaflets into audio for easier consumption.
With audio, there are tens of use cases that can have a strong social impact.

Armando Vieira, Founder @ Alea.ai
> Use of Generative Adversarial Networks for data imbalance problems

Abstract:
Generative Adversarial Networks (GANs) is a powerful approach to generate data from very few examples.
They have achieved state of the art results for image synthetization achieving photo-realistic quality in challenging problems like human faces.
In this presentation, I will show how GANs can be used for very imbalanced problems like insurance and fraud detection. The main idea is to use GANs to generate extra points in the minority class in order to make the data more balanced.
We will show how this compares with other methods like SMOTE and SMOTE+tomek.

Leonardo Alberto Dal Zovo, Chief Technology Officer @ Studiomapp
> How AI on Satellite Imagery supports Sustainable Development Goals

Abstract: The 17 Sustainable Development Goals (SDGs) are the world's best plan to build a better world for people and our planet by 2030. Adopted by all United Nations Member States in 2015, the SDGs are a call for action by all countries - poor, rich and middle-income - to promote prosperity while protecting the environment.
The increasing spatial and temporal resolutions of Earth Observation data offer an invaluable opportunity for better informing development policies and quantifying various SDG indicators.
In this talk we will present how Studiomapp is using Artificial Intelligence on Remote Sensing data to help in:

• SDG 11: support Disaster Response Management to make cities and human settlements more resilient
• SDG 14: battle illegal fishing by locating vessels that have not declared their position via AIS (a global vessel tracking service) to conserve and sustainably use the oceans, seas and marine resources
• SDG 9: monitor and assess the safety of existing infrastructure with low emissions.

Location: Microsoft Reactor

Agenda:

18:00 - 19:00 Networking and drinks

19:00 - 20:00 Talks followed by Q&As

20:00 - 21:00 Networking, drinks, and pizza sponsored by Logikk!

Who we are
At Logikk we engage exceptional humans for companies looking to unlock the potential of their data.
We believe people are at the centre of what we do, and we have taken that belief a step further by creating a community of data and machine learning professionals to enable networking, knowledge-sharing, and ideas spreading.

Find us here:
https://www.logikk.com/community/