Experience
Artera (November 2024 - )
Machine Learning Engineer
Artera is a leading precision medicine company developing AI tests to personalize cancer therapy.
Perceptive Space Systems (Jun 2024 - Jul 2024)
Machine Learning Scientist
Applied scientist building machine Learning-based systems for modeling the effects of space weather (radiation, magnetic fields) in low-orbit satellites. Responsible for data analysis (curation, quality assessment and processing) of years-worth of satellite data
Hinge Health (Jan 2022 - April 2024)
Computer Vision Scientist
CV Scientist I (Jan. 2022 – Aug. 2023), CV Scientist II (Sept. 2023 – April 2024): worked on multiple stages of Hinge Health’s Human Pose Estimation (HPE) pipeline. Examples include Auto-labelling (high-quality HPE), cloud-based model fine-tuning (AWS), data-driven scene parameters estimation (camera height, roll, pitch yaw), camera calibration, 3D-to-2D reprojection, lightweight HPE for processing-limited devices, 2D & 3D HPE.
Related media: Overview, CV at Hinge Health
University of Victoria ( Sept 2017 - Present)
Adjunct Professor - ECE Department (Sept. 2023 - Present)
Adjunct Professor at the Electrical and Computer Engineering department. Associate Member of the Faculty of Graduate Studies. Research interests are Computer Vision and adjacent areas.
Sessional Lecturer - Computer Vision (ECE 471/ECE 536) (Dec. 2020 - May 2021)
Sessional lecturer of the Computer Vision course (ECE 471/ECE 536) offered to undergraduate and graduate students at UVic in Spring 2021. Topics covered are an approximately equal amount of traditional Computer Vision and Deep Learning (see Teaching).
Traditional Computer Vision & Deep Learning Researcher (2018 - 2021)
Project 1 (Fall 2018 - Dec. 2021): Deep Learning-based Framework for the Detection of Pelagic Species in Multi-Frequency Echograms. In this project I apply deep learning- and traditional computer vision-based techniques to create autonomous detectors of schools of herring, salmon and hake, as well as swarms of krill in multi-frequency echograms. The echograms are based in acoustic backscatter data obtained by Canada's Department of Fisheries and Oceans (DFO) using AZFP echosounder instruments manufactured by ASL Environmental Sciences.
Project 2 (Jan. 2019 - Dec. 2021): Enhancement of Low-Lighting Underwater Images using Local Contrast and Multi-Scale Fusion. Local contrast information, physical priors and a multi-scale fusion process are used in this framework to enhance the visibility levels of underwater imagery. Current developments aim for the use of the proposed system in diverse layouts (e.g., out-of-water nighttime scenes).
Project 3 (Jan. 2019 - Aug. 2020): Marine Vessels Recognition in Short Time Series. Development of a hybrid system that uses traditional computer vision- and deep learning-based frameworks for the autonomous detection of marine vessels data that vary in size, shape, color, and level of visibility in environmental monitoring data. The vessels are to be detected in years-worth of photographs collected from two physical sites off the coast of Vancouver Island by the Geography Department of UVic.
Project 4 (Jan. 2021 - Dec. 2021): Autonomous Identification of Marine Vessels and Beluga Whales in the Canadian Arctic. Working with Environment and Climate Change Canada (ECCC) and the Wildlife Conservation Society Canada (WCS) to develop a system for the automatic detection of marine vessels and beluga whales using both visual and acoustic data gathered in-situ at multiple sites in the Canadian Arctic.
Related media:
Project 1: Magazine article, Web site article, presentation video, conference article
Project 2: Presentation video, Journal article, conference article
Project 3: Presentation video, conference article
Graduate Teaching Assistant (2017 - 2021)
Teaching assistant for the following courses: Design Project I, ECE 399 (Fall terms of 2017, 2018, 2019, 2020, 2021); Design Project II, ECE 499 (Summer term of 2021); Microprocessor-Based Systems, ECE 355 (Fall terms of 2017, 2018 and 2019); Medical Image Processing, ECE 435 (Spring term of 2020).
Related media: Teaching assistant award
Robotics Engineering Co-op at the Assistive Technology Laboratory (May 2018 - Sept. 2018)
Responsible for the navigation system of James, an autonomous robot created to assist physically and mentally challenged users. The navigation system uses the ROS (Robot Operating System) platform and its SLAM (simultaneous localization and mapping) and navigation tools.
Electrical Engineering Co-op at Ocean Networks Canada (Jan 2018 - May 2018)
California State University, Long Beach (May 2015 - Jul 2015)
Autonomous Robots Research Assistantship
Designed an autonomous robot capable of creating the map of a previously unknown environment. This mobile platform uses numerous sensors (i.e., Gyroscope, Ultrasonic and Quadrature Encores) and a Bluetooth Module controlled with an Arduino prototyping board.
Designed an application that creates a real-time map of the environment based on the mobile robot's measurements. This visual application was created using the Processing IDE.
Petrobrás, Brazil (Oct 2013 - Feb 2014)
Engineering Internship
Worked at the Research Center (CENPES) of Petrobrás. Activities performed involved the design and implementation of a lossless data compression and decompression algorithm using floating points. This novel system was Implemented using LabView and it was eventually employed as part of the control chain in oil refineries.
Federal University of Sergipe, Brazil (Aug 2009 - Aug 2012 / Nov 2016 - Apr 2017)
Sessional Lecturer - Robotics (Nov 2016 - April 2017)
60-hour Robotics course taught to Computer Science and Computer Engineering undergraduate students. Half of the course is devoted to the theoretical aspects of robotics (i.e., sensors, mechanics, control, locomotion, kinematics, etc.) while the other half is focused in real-world experiments using Arduino boards and multiple sensors, modules and "shields".
Research Internship (Aug 2009 - Aug 2012)
Developed projects involving 1) a rotating platform for a video camera using stepper motors as part of a surveillance system with Computer Vision-based face detection), 2) parallelization of Computer Vision algorithms, using both CPU (OpenMP) and GPU (CUDA), 3) an autonomous robot capable of locating itself (MCL algorithm), recognizing a target object (SURF features), and navigating to rescue it.
I have contributed as a reviewer to the following conferences and journals:
IEEE Transactions on Image Processing (2020, 2021)
IEEE Transactions on Circuits and Systems for Video Technology (2021)
IEEE Photonics Journal (2021)
WACV 2021, 2022 (main conference)
ICPR 2020 (main conference), 2018 (CVAUI workshop)
CVPR 2020 (NTIRE workshop)