Me

carlostorres[at]ece[dot]ucsb[dot]edu

carlos[dot]torres[dot]ee[at]gmail[dot]com

I got my PhD degree (link to defense deck) from the Department of Computer and Electrical Engineering (ECE) at the University of California Santa Barbara (UCSB) as a member of the Vision Research Laboratory under the advisement of Prof. B. S. Manjunath. My thesis focused on the design and deployment of distributed multimodal sensor networks (hardware, software, and analysis mechanisms) for human pose, action, activity, and event analysis in natural healthcare settings. I  also have an M.S. in Electrical and Computer Engineering from UCSB and a B.S. in Electrical Engineering and BioEngineering from San Jose State University.

I just recently joined TwoSixLabs in November as a Lead Research Scientist. My previous roles include Principal Investigator and Senior Researcher at Mayachitra Inc., where I developed new methods and solutions for activity and event analysis, uncertainty estimation, satellite imagery object detection, and smart multimodal distributed sensor networks; Data Scientist and Machine Learning Researcher at Procore Technologies in Carpinteria, where I designed and deployed methods to estimate and optimize construction-management processes; Lead Sr. Data Scientist at Carpe Data developing web-footprint based risk models for insurability applications (smoker, auto-loss, fraud, etc.); and as Computer Vision Engineer and Researcher at Caugnate (led by Matthew Turk, acquired by Vuforia in 2016). 

Resume & Curriculum Vitae (CV):

PhD Thesis Scope

My research focused on multimodal data analytics for healthcare applications by combining computational abilities with the healthcare system. In particular, I developed unobtrusive systems and statistical analysis methods and algorithms to autonomously monitor patients and healthcare environments. My research aimed to provide tools for objective identification and evaluation of medical therapies, protocols, and workflows along with their effect on patient health. The developed methods use statistical analysis and real-world datasets collected by an inexpensive multimodal multiview sensor network. These sensors continuously collected data in a medical ICU without disrupting patients or care practices.  The technical areas of my research involved coupled optimization, Hidden Markov Models, and conditional random fields.

Press Coverage

Spotlight Profile at UCSB-ECE 2015

MESH in the news Fall 2017

Research Interest

I am involved in multiple research projects in topics ranging from agriculture and water management and analysis to wound care, augmented reality, medical data analytics, and construction management. My attention is easily captivated by real-world data and problems. I seek to create robust autonomous systems that are adaptively aware of their surroundings. I am interested in the development of hardware and and software solutions (massive data!) that give systems the ability to autonomously analyze their surroundings and execute complex tasks in “natural” human environments. I am also interested in real-life problems where perfectly labeled data may not actually exist and where the impact levels are vast.