Miguel Saavedra-Ruiz

I am a Research Master's student with emphasis in Artificial Intelligence and Robotics advised by Liam Paull in Robotics and Embodied AI Lab (REAL) at Université de Montréal and MILA. Previously, I obtained a Postgraduate Diploma in Artificial Intelligence and a BEng degree as a Mechatronics Engineer from Universidad Autonoma de Occidente (UAO) in Cali, Colombia.

During the BEng degree, I worked under the supervision of Victor Romero-Cano on different areas such as object detection and tracking, mobile robotics, sensor fusion and deep neural networks. My undergraduate degree project was entitled "Autonomous landing system for an unmanned aerial vehicle on a terrestrial vehicle" and consisted in the development of the vision and control pipelines to autonomously land a UAV on a ground vehicle.

Over the past two years, my research has been focused on areas such as Reinforcement Learning, Bayesian Inference, Object Detection, and the application of AI in robotics-vision applications. Formerly I worked as a Machine Learning Engineer at Whale & Jaguar (NLP) and as an R&D Robotics Software Engineer at Romero Cano Ingenieria (Robotics vision, Multi-modal sensor fusion, AI, and UAVs.)

Email: miguel [dot] angel [dot] saavedra [dot] ruiz [at] umontreal [dot] ca

Github  /  Google Scholar  /  Linkedin  /  Email  /  CV

profile photo


I am broadly interested in the areas of Robotics vision, Machine Learning, Reinforcement Learning, Artificial Intelligence, Computer Vision, SLAM and Probabilistic Graphical Models. Most of my work is motivated by the following question: "How can robotic agents be endowed with environmental awareness to start perceiving their surroundings accurately and reliably and thus, enhance their decision-making capabilities?"


Monocular visual autonomous landing system for quadcopter drones using software in the loop
Miguel Saavedra-Ruiz, Ana Pinto, Victor Romero-Cano
IEEE Aerospace And Electronic Systems (to appear), 2021   (Journal publication)
code / poster / thesis / arxiv / video / description

Autonomous landing system for a UAV on a terrestrial vehicle using robotics vision and control.

3D object detector for vehicles using classic Machine Learning
Gustavo Salazar, Miguel Saavedra-Ruiz, Victor Romero-Cano
LatinX Workshop at CVPR, 2021   (Poster Presentation)
code / arXiv / poster / description

3D object detection of vehicles in the NuScenes dataset using classic Machine learning such as DBSCAN and SVMs.

Detection and tracking of a landing platform for aerial robotics applications
Miguel Saavedra-Ruiz, Ana Pinto, Victor Romero-Cano
CCRA, 2018   (Oral Presentation)
code / bibtex / video / description

Object Detection and tracking pipelines to detect a landing pad on the ground from a UAV.

Localization of a landing platform for a UAV
Miguel Saavedra-Ruiz, Ana Pinto, Victor Romero-Cano
(Poster Presentation)
code / poster

Localization of a landing pad located at the top of a ground vehicle with a UAV.

Style-transfer for the creation of aesthetic images
Miguel Saavedra-Ruiz, Gustavo Salazar, Sebastian Botero
code / report (spanish) / description

Style-transfer implementarion based on the paper A neural algorithm of artistic style using VGG-19.

Robotics Software Engineer projects
Miguel Saavedra-Ruiz
code / description

Robot localization, Mapping, SLAM, path planning and navigation.

Stereo Visual Odometry (VO) and Visual Inertial Odometry (VIO) with an Error-State Extended Kalman Filter in a quad-rotor
Miguel Saavedra-Ruiz
code / description

Stereo Visual Odometry and Visual Inertial Odometry pipelines to estimate the pose of a quad-rotor.

Simulation of a landing system for a UAV in Gazebo
Miguel Saavedra-Ruiz
code / video / description

Simulation of an autonomous landing system for a UAV with Gazebo, ROS and the Software in the loop provided by PX4.

​Reinforcement Learning Specialization Projects
Miguel Saavedra-Ruiz
code / video / description

Lunar Lander, Mountain Car and Pendulum classical control tasks solved using RL.

Teleoperation system for a car-like robot
Miguel Saavedra-Ruiz
code / video / description

Teleoperation system for a car-like robot through mathematical modelling.

Mapping and localization in indoors with Turtlebot 2
Miguel Saavedra-Ruiz
slides / report / description

Simulation of a localization and mapping system (laser-based SLAM) for a turtlebot2 in indoors.

​Self-Driving Cars Specialization Projects
Miguel Saavedra-Ruiz
video1 / video2 / description

Wheeled-robot mathematical modelling through dynamical modelling (tire model), lateral and longitudinal control, state estimation with Kalman filters, visual perception and motion planning for self-driving vehicles

​Low-cost license plate recognition system based on CNN
Miguel Saavedra-Ruiz

Implemented a low-cost license plate recognition systems using deep learning techniques.

Flow control with heatmaps in indoors
Miguel Saavedra-Ruiz

Heat map generator based on computer vision techniques to stochastically estimate the most visited areas in an indoor space with a monocular camera.

​Calendula Flower Classification System
Miguel Saavedra-Ruiz

Implementation of a low cost Calendula flower classification system using CNN.

Courses and Certifications
  • Reinforcement Learning by University of Alberta & Alberta Machine Intelligence Institute on Coursera. Certificate earned at June 21, 2020. [Credential]
  • Self-Driving Cars a 4-course specialization by University of Toronto on Coursera. Specialization Certificate earned on June 5, 2019. [Credential]
This template was shamelessly stolen from here.