Name: Malay Nagda
Type: User
Company: Arizona State University
Bio: Robotics- Artificial Intelligence graduate student at Arizona State University | Looking for Full Time position from Spring 2021 in AI/ML/Computer Vision.
Twitter: nagda_malay
Location: San Francisco, CA, United States
Malay Nagda's Projects
Implementation and comparison of D* Lite and A* search algorithms in the Pac-Man domain.
The RGB LED strip displays the average color of all the pixels of the laptop display with a music video playing.
The project involves 1. stereo depth sensing using camera calibration, stereo calibration and rectification; 2. camera motion and structure by calculating- a. homography to reconstruct the world plane from images; b. camera pose from a single view as a PnP problem; c. camera pose from two views using essential matrix.
Perform convolution as a matrix multiplication operation with a vertical edge filter.
Code for Harris corner detection, recover rotation and translation for a rotated and translated image.
A comparison of average comutation time for dense matrix multiplication in Scala without Breeze, Scala with Breeze and Apache Spark.
Implementation of face recognition, emotion detection, age and gender estimation in one system.
An attempt to see the result of swapping phases of two images and implement hybrid images that combine low frequencies of one image with high frequencies of other.
Generate Gaussian & Laplacian pyramids from scratch and tried multiresolution blending to have a seamless blend of images of an orange and an apple- an orapple.
Created a script to generate random squares and then trained a GAN to generate the same.
Multiple images are stitched together by feature detection, matching and warping the images with the help of homography matrix.
Made a 3x3x3 LED cube to code and display different interesting patterns using LEDs and Arduino.
A Term Project on Robotics Toolbox developed by Amaan Khan, Laukik Mujumdar, Malay Nagda, Ravendra Raghavendra and Tanmay Dhanote
The PACMAN agent finds path through a maze, both to reach a particular location and to collect food efficiently using algorithms like DFS, BFS, A* search and their different versions.
Various types of PACMAN agents like- Reflex agent, Adversarial agent (Minimax agent, Alpha-Beta Pruning and Expectimax agent) are designed.
An implementation of perceptron algorithm from scratch.
A project showcasing perception in robotics with ROS, Gazebo simulaor and OpenCV. It involves simulation for navigation, SLAM, object detection and tracking using turtlebot3 ROS package.
This static website was made as a capstone project for my undergraduate audit course of Web programming that included learning and implementing HTML, CSS and basic JavaScript.