Hi! I am Tanmay Randhavane. I am a graduate student of Computer Science at University of North Carolina at Chapel Hill. I am from a
in Maharashtra, India. I completed my Bachelor's studies in Computer Science at Indian Institute of Technology Bombay with honors and minor in Statistics. Currently, I am working under the guidance of Prof. Dinesh Manocha on a project involving multi-agent virtual interaction.
Pedestrian Behavior Learning | Ongoing
Developing algorithms to interactively learn pedestrian behaviors from various sources of crowd data.
Identify pedestrian personality traits and emotional states from the their trajectories, walking gaits, and other sources.
Using these behaviors to predict pedestrian trajectories in crowded situations, simulate pedestrian behaviors, and track pedestrians in crowd videos.
Design socially aware robot navigation algorithms that take into account learned pedestrian behaviors.
Experiencing Crowds in Virtual Reality | Ongoing
Developing algorithms to simulate interactive virtual crowds in a virtual environment along with a user.
Modeling the behavior of virtual agents to enable a user to approach the virtual agents and have face-to-face interactions.
Designing intelligent responses from the virtual agents to the attempts of interaction by the user agent.
Amazon Development Center, Bangalore, India | May 2014 - July 2014
Worked in the Amazon Fulfillment Technologies (AFT) team to create a testing framework for a data platform service .
Developed a user friendly framework providing the ability to create functional and integration tests, to generate mock messages and to publish them to corresponding queues.
Study of Mucus Hurricanes | Oct 2015 - Nov 2015
Mucus particles in human airway move in circular paths called "mucus hurricanes". Identified center of the circles from a set of images of mucus particles captured after regular time intervals.
Tracked particles in the set of images using an energy function based on location and intensity and fit circles to individual trajectories using Hough transform for circles. Compared the results with circles obtained by applying Hough transform to Maximal Intensity projection of the images.
Texture Modeling and Classification using Textons | Aug 2014 - Oct 2014
Worked on building a classification system of textures in images. Used texton based approach to model and represent visual textures.
A texton library was constructed from the extracted textons from training data. Each texture class was represented by a histogram on the texton library. Used chi square distance between histograms to classify textures into texture classes from training data.
Robustness Improvement of Speaker Verification | May 2013 - Jul 2013
Worked on improving an existing prototype system for real-time speaker verification against robustness issues.
Implemented an adaptive decision criterion driven by the signal-to-noise-ratio for an own voice speaker verification system.
Optimized and evaluated the new verification system for hearing aids in different noise environments.
Rendering using Ray Tracing and Point Cloud based Methods | Jan 2014 - Mar 2014
Implemented a Recursive Ray Tracer in C++ which reads a scene file and generates an image. Used Phong Illumination Model with point lights and uniform super-sampling for anti-aliasing.
Used Renderman by Pixar to generate images showing the effects of colorbleeding, caustics, area lights, soft shadows and textures using both ray tracing and point cloud based methods.
, Aniket Bera, Rohan Prinja, and Dinesh Manocha. "F2FCrowds: Planning Agent Movements to Enable Face-to-Face Interactions”,
To appear in Presence: Teleoperators and Virtual Environments 26-2, 2017.
, Rohan Prinja, and Dinesh Manocha. "SocioSense: Robot Navigation Amongst Pedestrians with Social and Psychological Constraints",
In IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2017.
, and Dinesh Manocha. "Aggressive, Tense or Shy? Identifying Personality Traits from Crowd Videos",
In International Joint Conference on Artificial Intelligence, pp. 112--118. IJCAI 2017.
Sahil Narang, Andrew Best,
, and Dinesh Manocha. "PedVR: Simulating Gaze-Based Interactions between a Real User and Virtual Crowds”,
In 22nd ACM Symposium on Virtual Reality Software and Technology, VRST 2016.
Aniket Bera, Sujeong Kim,
, Srihari Pratapa, Dinesh Manocha. "Realtime Pedestrian Path Prediction using Global and Local Movement Patterns",
In IEEE International Conference on Robotics and Automation, May 16-21, 2016, Stockholm, Sweden, ICRA 2016.
Aman Mangal, Arun Mathew,
, Umesh Bellur. "Predicting power needs in smart grids",
In Proceedings of the 8th ACM International Conference on Distributed Event-Based Systems, May 26-29, 2014, Mumbai, India, DEBS 2014.
SN 336, Sitterson Hall, Department of Computer Science, Chapel Hill, NC 27514.