Prochain Meetup

Thermal Face Authentication with Convolutional Neural Network - Author Presents
Today, Faris Baker, a regular in our group, will present his paper "Thermal Face Authentication with Convolutional Neural Network". Abstract: Matching thermal face images as a method of biometric authentication has gained increasing interest because of its advantage of tracking a target object at night and in total darkness. Therefore, for security purposes, it has become highly favourable and has extensive applications, for instance, in video surveillance at night. The aim of this study is to present a simple and efficient deep learning model, which accurately predicts person identification. ... The proposed approach is effective compared to the state-of-the-art thermal face recognition algorithms and achieves impressive accuracy of 99.6% with less processing and training times. Paper: https://thescipub.com/abstract/10.3844/ofsp.12216 Discussion: 1. Under what conditions is thermal better than visible light? 2. What are the entrepreneurial opportunities for this technology?

Algolia

55 Rue d'Amsterdam · Paris

À propos de ce groupe

The best way to become an expert in machine learning is to learn from the best. Tuesday meetups are BYOT (Bring Your Own Topic) i.e. open discussions. Each Friday meetup, we'll study a pre-selected research paper in machine learning. Papers may include image processing, AGI, NLP or other deep learning topics. The papers and our discussions are in English. Meetings are limited to 25 people so our peer-to-peer style discussions don't become overwhelming.

Hi, I'm Robert Salita, the Paris group's organizer. I'm a veteran of the successful San Francisco Deep Learning Study Group. They gave me two pieces of advise in starting the Paris group; meet regularly, weekly if possible, and develop a core group of talent. I'm particularly looking for ML math guys to join us. So many of the research papers have advanced math involved.

Don't worry about not fully understanding the research papers. Even top San Francisco members would comment that they only understood 75% of the paper. Papers are often short on key details, don't make source code available, or lack ways to experiment with or duplicate findings.

Membres (822)

Photos (2)