Kumiko Baland

I am currently working as a team manager of AI section in a surveying consulting company. I have worked as an engineer in the field of remote sensing / photogrametry / GIS. Some example of the recent projects that I joined in as a main developper are below:

• Developed a building change detection and web visualization tool brought by a customized algorithm for aerial images with a cross functional team, leading to human errors reduction and currently used in production lines in the company
• Implemented an estimation tool incorporated to GIS, which calculates sand particles’ size from close-up photos taken in boring surveys by AI and rule-based polygon unifying processing.
• Designed and implemented AI algorithm to estimate the depth from in-vehicle images and produced scripts to calibrate and project in-vehicle LiDAR point-clouds to verify the accuracy.
• Designed SLAM experiments with a stereo/panoramic camera, and implemented a script to validate the accuracy via point cloud processing.
• Raised research funds through competitive tender and designed/implemented forest cover classification algorithms using multi-modal data.

2021/04 – Present Senior Engineer/Team Leader @AI unit, Kokusai Kogyo Co., Ltd. 
• Supervised AI-related initiatives by setting appropriate goals and working together with external departments on over 10 PoC / R&D projects while managing a team of 4 project leaders, resulting in 20 % operational cost reduction at most.
• Acquired and supported teams to acquire several patents regarding AI systems and pre-/post-processing methods through actively reviewing projects
• Developed a building change detection and web visualization tool brought by a customized algorithm for aerial images with a cross functional team, leading to human errors reduction of 5% and currently used in production lines in the company

2015/10 – 2020/11 Senior Engineer @AI unit, Kokusai Kogyo Co., Ltd. 
• Launched an airborne laser filtering project composed with spanning multiple teams including subsidiary, which detect the area to be reviewed by operators using an AI/point-clouds processing, providing up to 20 % reduction in human-resources and used in production lines.
• Collected and annotated training data and implemented a tool to extract tree crown polygons using object detection AI from drone images.
• Designed and implemented AI algorithm to estimate the depth from in-vehicle images and produced scripts to calibrate and project in-vehicle LiDAR point-clouds to verify the accuracy.
• Designed SLAM experiments with a stereo/panoramic camera, and implemented a script to validate the accuracy via point cloud processing.
• Raised research funds through competitive tender and designed/implemented forest cover classification algorithms using multi-modal data.
 

2012/04 – 2014/08 Project Engineer @Remote Sensing unit Kokusai Kogyo Co., Ltd.
• Implemented a script to estimate wheat yield using hyperspectral images by performing dimensional reduction.
• Developed a simulator of multi agent system for a municipal to estimate evacuation status.
• Developed a tool incorporated to GIS that converts satellite images from digital numbers to reflectances by reading variables from metadata and applying a conversion formula.
 

2018/06 – 2020/10 Part-Time Engineer @National Institute of Advanced Industrial Science and Technology
• Assisted model implementation and executed exercises for ablation study, where changes along roadside are detected by applying a self-supervised AI-based image alignment method to in-vehicle images, of which results showed effectiveness compared to existing models, leading to acceptance in the IROS conference .

2014/09 – 2015/12 MSc. Artificial Intelligence KU Leuven  Major in Big Data Analytics. Leuven, Belgium
2010/04 – 2012/03 MSe. Social Engineering Tokyo Institute of Technology  Major in Geographic Information Science and Cartography. Tokyo, Japan
2006/04 – 2010/03 Ba. Economics Keio University  Major in Regional Analysis. Tokyo, Japan