Hello! I am Mrityunjay Mishra, a rising senior at the University of Texas at Austin. I am pursuing a bachelor in Computer Science, along with a minor in Business. I am also an undergraduate researcher with the Learning Agents Research Group (LARG) at UT Austin, and I am currently working under Dr. Justin W. Hart. My research interests lie in the are of human-robot interaction (HRI), and I am currently researching how different gaze patterns can affect the interpretability of the target of an agent's gaze.
In addition to school, I also work as a part-time software consultant. I have developed a test automation software for F2G-Solutions, and I have also developed a 5G-testing toolkit (using the ThingSpace API by Verizon) for The 5G-Garage Inc. Currently, I am developing an API to automatically generate CRUD APIs for Pella Corporation.
Outside of school and work, I like to play tennis. I also love to go bouldering, running, and working out in general. When I am not working out, I am usually hanging out with my friends. Besides that, I love playing single-player video games. I also love cooking, reading, and watching TV.
In the collatz project, I, along with a partner, implemented a basic python program that printed the cycle length based on the collatz conjecture for my Software Engineering class (SWE). The main purpose of this project was to implement a program that calculates the cycle length based on the collatz conjecture and to also learn about other useful tools. I learned about source control using git and GitLab; I also learned how to establish our own CI/CD pipeline in GitLab. Furthermore, I learned how to open, track, and close issues in GitLab. My partner and I wrote unit tests for our program using the unittest class in python, and we also verified the reachability of our unit tests through coverage. We also checked our program through public acceptance tests and Sphere (SPOJ).
I wrote an inverse-kinematic (IK) solver using the Ceres library by Google as part of a larger research project for the Learning Agents Research Group. In this project, I wrote an IK solver in C++ using the Ceres library by Google to control the motion of a Maki Humanoid Robot Head. To implement the solver, I first wrote a forward-kinematic solver that provided a forward-kinematic map to ceres, which used non-linear least squares optimization (using the forward-kinematic map in the residual function) to output the joint values that would minimize the distance between the current joint configuration and the desired joint configuration. Once written, I intergrated this IK solver into the larger experiment being conducted at LARG.
+1 (940) 367-3183