Edgar Medina Email & Phone Number

Munich, By, De

Munich, Bavaria, Germany

Research in the autonomous driving area, in particular:- Creating new interpretable architectures for 3D Human motion prediction using GCNs.- Research and implementation of NeRF architectures for novel view synthesis and model robustness tests.- Investigating various adversarial attack techniques within GCN frameworks to assess robustness.- Evaluating cutting-edge architectures to benchmark performance in predicting 3D human motion tasks.- Developing reinforcement and… Show more Research in the autonomous driving area, in particular:- Creating new interpretable architectures for 3D Human motion prediction using GCNs.- Research and implementation of NeRF architectures for novel view synthesis and model robustness tests.- Investigating various adversarial attack techniques within GCN frameworks to assess robustness.- Evaluating cutting-edge architectures to benchmark performance in predicting 3D human motion tasks.- Developing reinforcement and imitation learning algorithms to train humanoids to perform ‘walking’ and ‘running’actions.- Code deployment using Dockers, Jenkins, and Kubernetes for training and inference.list of publications: 3 papers in (1) WACV and (2) VISAPP.Research projects: KI-Absicherung, KI-Atenttion Show less

Darmstadt, Hessen, Deutschland

- Drive and contribute to the development of new image & video analytics algorithms.- Conduct research and coordinate work with peers, external partners and customers to develop a content base.

Stuttgart Area, Germany

- 3D processing data using computer vision and machine learning for Digital Twin systems.- Uncertainty analysis using Bayesian networks and model interpretation for decision-making explanations.- Robotics using deep Q-learning and policy gradient methods.- 3D object identification and segmentation using deep networks on point cloud and multi-view algorithms.- Synthetic data generation to train neural networks for object detection and segmentation tasks (GANs and/or… Show more - 3D processing data using computer vision and machine learning for Digital Twin systems.- Uncertainty analysis using Bayesian networks and model interpretation for decision-making explanations.- Robotics using deep Q-learning and policy gradient methods.- 3D object identification and segmentation using deep networks on point cloud and multi-view algorithms.- Synthetic data generation to train neural networks for object detection and segmentation tasks (GANs and/or Blender).- Pose estimation, real-time tracking, and action recognition in videos for safety applications. Show less

Curitiba Area, Brazil

Generate documentation, codes, and reports about my work and outcomes. Also, share, discuss and present findings with our partners. Work alongside engineers on topics related to computer vision and deep learning for real-time object detection systems.Main activities are the following:- Lead projects in the area of data analytics for the retail industry. Work alongside computer vision engineers and software developers to transform algorithms into commercially feasible products and… Show more Generate documentation, codes, and reports about my work and outcomes. Also, share, discuss and present findings with our partners. Work alongside engineers on topics related to computer vision and deep learning for real-time object detection systems.Main activities are the following:- Lead projects in the area of data analytics for the retail industry. Work alongside computer vision engineers and software developers to transform algorithms into commercially feasible products and services.- Design and implement the entire object detection system using cutting-edge learning-based techniques.- Develop computer vision algorithms for analytics aiming to understand the behavior of customers in retail industry. Show less

Rio De Janeiro, Rio De Janeiro, Brazil

Working together with GE researchers as consultant in topics related to computer vision and deep learning for 2D automatic segmentation systems. Employing mainly mathematical methods and deep neural networks. In particular, fully convolutional networks (FCN) were used for image segmentation and generative adversarial networks (GANs) were used for image completion.Libraries employed are Pytorch, Keras for deep learning; and OpenCV for computer vision algorithms

Rio De Janeiro Area, Brazil

Member of the team responsible of working with medical imaging. Works related to computer vision, deep learning techniques and image processing mainly segmentation.Project links: - https://www.youtube.com/watch?v=tJGuPvSxWfw- http://newsroom.gehealthcare.com/this-ai-tech-is-changing-mr-neuro-imaging/Main activities are the following:- Image registration techniques for data augmentation on 2D medical images.- Object orientation estimation in large structures… Show more Member of the team responsible of working with medical imaging. Works related to computer vision, deep learning techniques and image processing mainly segmentation.Project links: - https://www.youtube.com/watch?v=tJGuPvSxWfw- http://newsroom.gehealthcare.com/this-ai-tech-is-changing-mr-neuro-imaging/Main activities are the following:- Image registration techniques for data augmentation on 2D medical images.- Object orientation estimation in large structures employing CNN and FCNN.- 3D object detection in fine structures based on CNN for medical image segmentation.Libraries employed were Tensorflow, Keras, Pytorch for deep learning; and OpenCV, SimpleITK for computer vision algorithms. Show less

Lima, Peru

Working for the Direction of Scientific Instrumentation (DINCI) as Practitioner at the Space Agency of Peru (CONIDA). Working with the design and elaboration of the payload for sounding rockets and space instruments for scientific applications.Libraries employed were ANSI C, C, C++ for microcontroller programming; and MATLAB, Python for algorithm analysis and signal processing.

Master’S Degree, Computational Intelligence, Electronic Engineering

Activities and Societies: Works related to machine learning techniques, computer vision and image processing. Active member from Analog.

Bachelor’S Degree, Ingeniería Electrónica, Ingeniería Electrónica

Student and active member at Development and Research Center for Engineering (CIDI), Works related to Microcontrollers and FPGAs.