We created a GUI for user to capture images. We then perform data augmentation on it and followed by training the model using ResNet152. It also had a detection system, inventory management interface and a 2DoF Robotic arm simulation.
110 kV Switchyard Maintenance and Monitoring Robot. Developed a robot for cleaning of post insulators, greasing of isolators and automated surveillance of the switchyard.
Recommending and Monitoring Physiotherapy exercises. We created an application that does pose estimation using PoseNet model, then used our custom Classifier model to check if the user is in the required pose, the repetitions, accuracy and the next move.
Built a warehouse simulation using Gazebo where packages will be managed using two robotics arms, two logical cameras, one color camera to read QR codes and a speed variable conveyor belt. Built a web-based monitoring layer on top of it that displays information about the packages that were processed.
This tool takes the input as the playlist link and responds consists of the number of videos, the average duration of each video, and the duration of all the videos in varying playback speeds(1x, 1.25x, 1.5x, 1.75x, 2x).
Can query on the Pokemon database by asking questions in direct text. Uses Google AI PaLM 2 model text-bison and Langchain to understand your question and creates a SQL query. Runs that query in Azure SQL database and pressent the answer.