Hu Wenmiao

Hu Wenmiao

Ph.D. candidate & Ph.D. Researcher

Biography

Senior Data Scientist at Grab.

I am currently a Senior Data Scientist at Grab, and hold a Ph.D. degree from National University of Singapore. My Ph.D. thesis “Usage of very high-resolution optical RGB satellite imagery in geo-information extraction for fine-scale map making” studied different methods and sub-topics to use satellite imagery in geoinformation extraction for map making with fusion with other data sources. In the Ph.D. we explored methods related to synthetic image generation, image segmentation, scene understanding, cross-view matching, etc.

Before this, I obtained my Master of Science degree (2016) at Technical University Munich (Germany), Earth Oriented Space Science and Technology (ESPACE), where I gained knowledge about earth system science, remote sensing, and navigation. My master thesis ”Object-based multi-view façade matching in SAR images of dense urban areas” focuses on matching building façades in Synthetic Aperture Radar data from multiple different viewing angles via radargrammetry. I received a Bachelor of Engineering degree from Nanyang Technology University (Singapore), School of Electrical and Electronic Engineering in 2014.

My research interests include location-based augmented/mixed reality, satellite imagery, geo-localization and scene understanding.

Download my resumé

Education
  • Ph.D., Computer Science, completion in 2024

    National University of Singapore

  • Master's Degree (Passed with high distinction), Earth Oriented Space Science and Technology (ESPACE), 2016

    Technische Universität München, Munich, Germany

  • Bachelor's Degree (First Class Honours), Electrical and Electronics Engineering, 2014

    Nanyang Technological University, Singapore

Skills

Deep Learning
GeoInformation Extraction
Image Generation
Computer Vision
Satellite Image
Street-view Image
Maps

Experience

 
 
 
 
 
Senior Data Scientist
Grabtaxi Holding Pte. Ltd.
Oct 2024 – Present Singapore
  • Senior data scientist in Geo-Data Science team.
  • Conducting research on sensor fusion and localization.
 
 
 
 
 
Ph.D. Researcher
Grabtaxi Holding Pte. Ltd.
Aug 2019 – Sep 2024 Singapore

Full-time joint position with Ph.D. work.

Responsibilities include:

  • Collaboration with Geo-Data Science team.
  • Conducting research on geo-information extraction from imagery.
  • Co-filed five patents (three granted).
 
 
 
 
 
Product Manager and Solution Developer
Skymap Global Pte. Ltd.
Nov 2016 – Feb 2019 Singapore

Responsibilities include:

  • Conducted market research, product design, algorithm development for satellite imagery based insight generation and services.
  • Applied for and was awarded a government grant.
  • Managed multi-cultural product development group and recruited team members.
 
 
 
 
 
Research Assistant
German Aerospace Center (DLR) / TUM Signal Processing in Earth Observation
Mar 2016 – Sep 2016 Oberpfaffenhofen, Germany
Master thesis: Object-based multi-view façade matching in SAR images of dense urban areas.
 
 
 
 
 
Research Assistant
TUM Institute of Astronautics
Mar 2016 – Sep 2016 Munich, Germany

Tasks:

  • Supported data collection and performed data analysis for projects.
  • Vehicle route detection and visualization from GPS logs.
 
 
 
 
 
Internship
Panasonic R&D Center Singapore
Mar 2013 – Aug 2013 Singapore

Tasks:

  • Developed algorithm for nail beautification application.
  • Improved car plate recognition algorithm.

Education

 
 
 
 
 
Ph.D. (Computer Science)
National University of Singapore
Aug 2019 – Present Singapore

Thesis: Usage of very high-resolution optical RGB satellite imagery in geo-information extraction for fine-scale map-making.

Supervisor: Prof. Roger Zimmermann

 
 
 
 
 
Master of Science (Earth Oriented Space Science and Technology)
Technische Universität München
Oct 2014 – Sep 2016 Munich, Germany

Major in Remote Sensing & Synthetic Aperture Radar (SAR). Passed with High Distinction.

Thesis: Object-based multi-view façade matching in SAR images of dense urban areas.

Supervisor: Prof. Xiaoxiang Zhu.

 
 
 
 
 
Bachelor of Engineering (Electrical and Electronic Engineering)
Nanyang Technological University
Aug 2010 – Aug 2014 Singapore

First Class Honours

FYP: Human Emotion Recognition to Improve Driving Safety.

 
 
 
 
 
Summer Study
University of California
Aug 2011 – Aug 2011 Berkeley, CA, USA
Two modules on marine science and human resources.

Collaborations

(2023). Multimodal Deep Learning for Robust Road Attribute Detection. ACM Transactions on Spatial Algorithms and Systems.

Cite DOI

(2023). CrossMatch: Source-Free Domain Adaptive Semantic Segmentation via Cross-Modal Consistency Training. In ICCV 2023.

PDF Cite

(2021). Multimodal Fusion of Satellite Images and Crowdsourced GPS Traces for Robust Road Attribute Detection. In SIGSPATIAL ‘21.

PDF Cite DOI