Skip to main content

Welcome to AI and GIS production in NLS -seminar 1.12.2021

AI and GIS production in National Land Survey of Finland -winter seminar is going to be held on 1.12.2021. Welcome all of you who is interested in the AI development in geospatial field! Registration is open until 28.11.

Visualisation of building vectors from above
Photo:
National Land Survey of Finland

In “Advanced Technology for National Topographic Map Updating” project (ATMU) National Land Survey of Finland investigates how AI (deep learning technology) can be used to update topographic data.

After the Spring Seminar, the ATMU-project has made great progress. In the past half a year, we had a significant movement from the AI beginner to the AI technology developer. We have dived deeply in studying the possibilities of the AI for National topographic map updating. 

Welcome to our winter seminar, which will be held on 1.12.2021 at 13.00-15.00. In this seminar, we will share our knowledge with you. 

Seminar language is English and the event will be held on Teams. 

Registration

Register before 28.11: https://www.lyyti.fi/reg/AIGIS2021. You will receive the Teams link after you have registered to the event.  

For more information, please contact Lingli Zhu or Johanna Ujainen, firstname.surname@nls.fi

About the ATMU project in general, please visit the webpage (in Finnish).

Winter seminar programme

13.00-13.50          

  • Introduction 
  • Keynote talk: Artificial Intelligence in the Era of Big Data and Convergence Science. Keynote speaker: Dr. Zhang Zhe, assistant professor, from Texas A&M University, USA.  
  • Introduction to the status of the ATMU project  
  • Using the U-Net for building detection and outline extraction

13.50-14.00 Break
  
14.00-15.00

  • Using deep learning for the hydrographic feature detection: the status report 
  • Using transfer learning technology for building detection 
  • Using multitask learning for road detection 
  • Deep learning in Remote Sensing and GIS  
  • Discussion and conclusions
     
Spatial data
Research
Innovations

News and articles