Tertiary Institutes Leverage
Research Capabilities Through GIS
With AI and GIS serving as the driving forces behind urban innovation, cities can transform into dynamic and responsive entities that improve people’s daily lives. Organized by the Engineering Discipline of the Hong Kong Institute of Vocational Education (IVE Engineering), the hybrid seminar “Empowering the Metropolitan: How AI is Shaping Smarter Cities of Tomorrow” provided a platform for technology sharing on how AI can provide a more dynamic, data-driven, and citizen-centric process for the development of smart cities. Sr Paul Tsui, Chief Executive Officer of Esri China (Hong Kong), delivered a presentation titled “GeoAI Illuminates the Path to Future Smart”, showcasing the capabilities of GeoAI through real-world applications, including the Segment Anything Model (SAM) that can automate the feature extraction process across large geographical areas.
Sr Paul Tsui, Chief Executive Officer of Esri China (Hong Kong), showcased the “TextSAM” deep learning model for extracting spatial features using text prompts in ArcGIS Pro.
Apart from IVE Engineering, the JC STEM Lab of Earth Observations and Research Centre for Artificial Intelligence in Geomatics (RCAIG) at the Hong Kong Polytechnic University (PolyU) has transformed urban environmental research through theoretical and technical advancements. The Research Center recently invited Sr Paul Tsui to give a presentation on the topic of “GeoAI - A New Era of Location Intelligence for Urban Management”.
Sr Paul Tsui, Chief Executive Officer of Esri China (Hong Kong), interacted with the audience in the seminar at the Hong Kong Polytechnic University.
The capabilities of GeoAI have the potential to significantly enhance geospatial workflows and speed up the extraction of meaningful insight through satellite imagery, text data, time-series data and more. Sr Paul Tsui has shared the library of over 70 pretrained GeoAI models from the ArcGIS Living Atlas so that anyone can use it as it is, or as a starting point when creating their own models.