Lessons Learned Training Object Detection Models on Satellite Imagery

Описание к видео Lessons Learned Training Object Detection Models on Satellite Imagery

Roger FONG Computer Vision Scientists at Picterra
SwissAI Machine Learning Meetup 2018.10.15

Some of the addressed themes:

1. How Can We Use Satellite and Drone Images?

2. What Models And Datasets Can We Use to Learn How to Label Objects On Satellite Images?

3. What Is YOLO - You Only Look Once, Real-Time Object Detection Model

4. What Are The Solutions to Common Problems When Analysing Satellite Images with YOLO?

Abstract: Satellite imagery provides a wealth of visual information but without the proper tools, it can be daunting to understand. Our mission at Picterra is to provide deep learning empowered tools that are capable of analyzing these images and providing valuable information to our customers. In this talk, we will focus on the challenges of object detection from satellite imagery and our approach to the problem. https://picterra.ch

Organizers ##
SwissAI Machine Learning Meetup is one of the larges AI meetups in Switzerland, with regular meetings, great speakers invited from academia and industry and over 1200 members from Lake Geneva Area. For more information and future events visit https://www.SwissAI.org

Pawel Rosikiewicz, Founder of SwissAI,, Event Organiser
  / pawel-rosikiewicz  

Juraj Korček, Data Scientist and ML Engineer, Event co-organiser and Interviews
  / korcekjuraj  

Ieva Vaišnoraitė-Navikienė, ML Engineer, Event co-organiser
  / ieva-vaisnoraite-navikiene  

Matteo Pagliardini, Senior ML Engineer, Event co-organiser
  / matteo-pagliardini  

Clement Charollais, EPFL, Camera Operator and Movie Editing
  / clément-charollais-b7209177  


Sponsors:
École Polytechnique Fédérale de Lausanne (EPFL)
https://www.epfl.ch

Innovaoud
https://www.innovaud.ch

SamurAI - Data Science Services
https://www.samurai.team

Комментарии

Информация по комментариям в разработке