Omid Karimpour

• 

  Autonomous Mobile Robot Project

  Developed and implemented a modified Pioneer 2DX Autonomous Mobile Robot using RPLidar, localization, SLAM, mapping, path planning, and obstacle avoidance.

  Various SLAM methods were surveyed such as Gmapping, EKF, and Hector. Investigated challenges like Stair mapping, glass detection, and navigation

  You can watch a demo of my work here

• 

  Deep Learning Project on DNN and CNN

  Built a L-layered Neural Network using Vectorization, different Regularization, multi classification and Utilized various optimization techniques such as Adam, RMSprop, Momentum optimization, Regularization, Dropout, and cut off.

  Built and Applied Convolutional Neural Network such as eNet-5, AlexNet, VGG-16 on several data sets with Object localization, Landmark Detection, Object Detection

• 

  Overlay Image Project using OpenCV

  Utilized Amazon EC2, Lambda Function and S3 Storage services for Continuous Deployment and Included the continuous Integration using Travis and GitHub for new versions of the code and a complete instruction

• 

  Object Recognition

  Digit recognition using prediction models, car acceptability, wine quality, & water level. Further, Modeled and simulated with MATLAB on different real world datasets

  Created a Neural Network, SVM Prediction Model, K-means clustering for image dimensionality reduction

• 

  Object Detection

  Lines, & circles detection in noisy environment with 5 pixels accuracy. In addition, Blurring Process, Edge detection, and Hough transform algorithms with OpenCV

  Comparison of Hough transform; Fast and Random Hough transform; and efficient randomized algorithm

  Details

Gallery

Sometimes a picture needs no words, Photography is the science, art and practice of creating durable images by recording light.