• Curriculum Vitae
  • Google Scholar
  • Linkedin

About me

Welcome to my homepage! I am a doctoral research student at the Department of Computer Science in the University of Maryland, College Park, USA. My research interests lie in Computer Vision, Computer Graphics and Robotics. I am currently working with Dr. Dinesh Manocha to develop algorithms for human emotion classification from faces, speech and gaits, as well as generate skeletal models of human gaits corresponding to various emotions. Before coming to Maryland, I completed my Master of Engineering and then worked one year as a project associate under Dr. Venu Madhav Govindu at the Department of Electrical Engineering in the Indian Institute of Science, developing novel algorithms for robust 3D reconstructions of objects and scenes from raw depth images.

Research

DensePeds: Pedestrian Tracking in Dense Crowds Using Front-RVO andSparse Features
Rohan Chandra, Uttaran Bhattacharya, Aniket Bera, Dinesh Manocha. IROS 2019 Oral. [Dataset] [Project] [PDF] [BibTeX]

A pedestrian tracking algorithm that tracks individuals in highly dense crowds (>2 pedestrians per square meter). Our approach is designed for videos captured from front-facing or elevated cameras. We present a new motion model called Front-RVO (FRVO) for predicting pedestrian movements in dense situations using collision avoidance constraints and combine it with state-of-the-art Mask R-CNN to compute sparse feature vectors that reduce the loss of pedestrian tracks (false negatives). We evaluate DensePeds on the standard MOT benchmarks as well as a new dense crowd dataset. In practice, our approach is 4.5x faster than prior tracking algorithms on the MOT benchmark and we are state-of-the-art in dense crowd videos by over 2.6% on the absolute scale on average.

TraPHic: Trajectory Prediction in Dense and Heterogeneous Traffic Using Weighted Interactions
Rohan Chandra, Uttaran Bhattacharya, Aniket Bera, Dinesh Manocha. CVPR 2019 Poster. [Project] [PDF] [BibTeX]

A novel algorithm for predicting the near-term trajectories of road-agents in dense (>3000 road agent per Km) and heterogeneous (>5 different types of agents simultaneously present) traffic videos. The road-agents may correspond to buses, cars, scooters, bicycles, or pedestrians. We model the interactions between different road-agents using a novel LSTM-CNN hybrid network for trajectory prediction. In particular, we take into account heterogeneous interactions that implicitly accounts for the varying shapes, dynamics, and behaviors of different road agents. In addition, we model horizon-based interactions which are used to implicitly model the driving behavior of each road-agent. We evaluate the performance of our prediction algorithm, TraPHic, on the standard datasets and also introduce a new dense, heterogeneous traffic dataset corresponding to urban Asian videos and agent trajectories. We outperform state-of-the-art methods on dense traffic datasets by up to 30%.

Efficient and Robust Registration on the 3D Special Euclidean Group
Uttaran Bhattacharya, Venu Madhav Govindu. ICCV 2019 Poster. [PDF] [BibTeX]

An accurate, robust and fast method for registration of 3D scans. Our motion estimation optimizes a robust cost function on the intrinsic representation of rigid motions, i.e. the Special Euclidean group SE(3). We exploit the geometric properties of Lie groups as well as the robustness afforded by an iteratively reweighted least squares optimization. We also generalize our approach to a joint multiview method that simultaneously solves for the registration of a set of scans. We demonstrate the efficacy of our approach by thorough experimental validation. Our approach significantly outperforms the state-of-the-art robust 3D registration method based on a line process in terms of both speed and accuracy. We also show that this line process method is a special case of our principled geometric solution. Finally, we also present scenarios where global registration based on feature correspondences fails but multiview ICP based on our robust motion estimation is successful.

Fast Multiview 3D Scan Registration Using Planar Structures
Uttaran Bhattacharya, Sumit Veerawal, Venu Madhav Govindu. 3DV 2017 Spotlight. [PDF] [BibTeX]

A fast and lightweight method for 3D registration of scenes by exploiting the presence of planar regions. Since planes can be easily and accurately represented by parametric models, we can both efficiently and accurately solve for the motion between pairs of 3D scans. Additionally, our method can also utilize the available non-planar regions if necessary to resolve motion ambiguities. The result is a fast and accurate method for 3D scan registration that can also be easily utilized in a multiview registration framework based on motion averaging. We present extensive results on datasets containing planar regions to demonstrate that our method yields results comparable in accuracy with the state-of-the-art while only taking a fraction of computation time compared with conventional approaches that are based on motion estimates through 3D point correspondences.

Education

University of Maryland, College Park 2018 - Present

Ph.D. in Computer Science.       Advisor:   Dr. Dinesh Manocha

Selected Coursework:

Spring 2019:


Fall 2018:


Teaching Assistant:

Fall 2018:


GPA:   4.0/4.0   (After Spring 2019)

Indian Institute of Science 2015 - 2017

M.E. in System Science and Automation.       Advisor:   Dr. Venu Madhav Govindu

Selected Coursework:

Fall 2016:

Spring 2016:

Fall 2015:


Teaching Assistant:

Fall 2016:


GPA:   6.8/8.0 (≡ 3.8/4.0)

West Bengal University of Technology 2011 - 2015

B.Tech. in Computer Science and Engineering

From the Institute of Engineering and Management under WBUT.
GPA:   9.31/10.00 (≡ 3.9/4.0)

Awards

  • Dean's Fellowship, University of Maryland, 2018
  • Outstanding Student Award, Institute of Engineering and Management, 2013




      Last updated: August 27, 2019