Daile Osorio-Roig

Data scientist

Development of website scraping using Python Scrapy framework in order to collect occupancy and inspection data of the buildings on different United States regions for Fire Department planning.

Development of microservices - based architectures on .Net to exchange messages (test orders and test results) between Laboratory Instrument Systems (LIS) and numerous medical instruments. Specifically, I have implemented new algorithms to parse ASTM, Kermit and HL7 message protocols. Finally, implemented several MVC functions on WinForms – based LIS desktop application.

Researched Computer Vision, Machine Learning and Pattern Recognition techniques applied to Biometric applications. Specifically, developed a new fusion method at segmentation level based on clustering to improve the iris segmentation. Moreover, proposed new multi-class deep learning approaches for iris segmentation and glass detection tasks, which improve traditional state-of-the-art bi-class techniques. In this context, a new Multi-Class Iris Segmentation database (MCIS) was created.

Developed Computer Vision, Machine Learning, Pattern Recognition and Image processing algorithms. More specifically, proposed new unsupervised classification methods for agricultural crop in UAV RGB images. In addition, developed new techniques for automatic car number plate recognition, proposed a new framework based on image quality evaluation for Biometric Video Iris Recognition in visible spectrum and implemented a new iris segmentation approach based on hand-crafted features.