Joaquin Gabaldon

Postdoctoral Research Fellow

University of Michigan•  August 2021 - January 2023

* Extended development of robotics methods for monitoring marine mammals through computer vision, deep learning, and mechatronics. * Explored underwater speed estimation through accelerometer vibration analyses in a series of physical experiments. * Mentored graduate students on wearable sensor data analysis and experimental deployment. * Built a website with a multi-institute team to educate researchers in using and analyzing animal-borne biologging tags and their data.

Graduate Student Research Assistant

University of Michigan•  September 2015 - July 2021

* Implemented Deep Neural-Network Faster R-CNN object detection for marine animal tracking, enabling a longitudinal study on dolphin habitat use and kinematics. * Built, calibrated, and validated an on-animal speed sensor for marine mammals. * Fused animal-borne sensor odometry with vision-based NN position detection to identify and track an individual within a group in a dynamic environment. * Developed a framework to estimate marine mammal energetics and efficiencies using an on-animal speed sensor. This was applied in a first-of-its-kind study persistently monitoring animal propulsive effort and behavior during varying-intensity tasks.

University of Michigan - Ann Arbor, Ann Arbor

Robotics, PhD•  September 2015 - July 2021

Profs. Kira Barton and K. Alex Shorter Topics: Marine mammal localization/metabolics/behavior analysis, computer vision, deep-learning, sensor fusion, pose estimation. Dissertation: Contextualized Monitoring in the Marine Environment