Abhinav Kumar

Publications / Google Scholar / DBLP / OpenReview / Semantic Scholar / Linkedin / Github / Twitter / Facebook / Quora / Resume

  • Email abhinav3663[at]gmail.com, kumarab6[at]msu.edu

  • Mobile +1 801 635 8185

  • Home Address 3636 Coleman Rd, #724C, Block-36 Apartments, East Lansing, MI 48823

Welcome! My name is Abhinav Kumar, and I am a PhD student in the Computer Vision Lab, Department of Computer Science at Michigan State University (MSU), working with Prof Xiaoming Liu. I transferred from the School of Computing, University of Utah. My research interests are applying Computer Vision and Mathematical Modelling to 3D Object Detection.

I am currently interning in the Computer Vision group at Facebook AI Research (FAIR), Menlo Park, CA, with Dr Georgia Gkioxari and Dr Garrick Brazil on 3D Object Detection. I spent Summer 2019 in the Computer Vision group at Mitsubishi Electric Research Labs (MERL), Cambridge, MA, working with Dr Tim Marks, Dr Michael Jones, and Dr Anoop Cherian on uncertainty and visibility estimation using deep neural networks.

During 2016-2018, I worked as a Research Engineer in Computer Vision group at Xerox Research Centre India/Conduent Labs India, Bangalore, with Dr Sriganesh Madhvanath and Vladimir Kozitksy on large scale Image Re-Identification of License Plates (LPR) and Dr Manasa Kolla on Imbalanced Classification for Vehicle Passenger Detection System (VPDS). Xerox/Conduent deploy my works on LPR and VPDS and use them in real life.

In my previous life, I was a Signal Processing and Communications Masters student guided by Prof Animesh Kumar in the Electrical Engineering department at Indian Institute of Technology Bombay (IIT Bombay) and a Bachelors student in the Electrical Engineering department at Indian Institute of Technology Patna (IIT Patna). I attended MK DAV Public School, Chianki, Daltonganj till 2007 and Jawahar Vidya Mandir, Shyamali, Ranchi for +2.

Publications

Visit Google Scholar / DBLP for latest ones and for more details. * = Equal Contributions

Conferences

GrooMeD-NMS: Grouped Mathematically Differentiable NMS for Monocular 3D Object Detection

Abhinav Kumar, Garrick Brazil and Xiaoming Liu
CVPR 2021
pdf / supplementary / 5min talk / slides / poster / demo / arxiv / code / project

LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood

Abhinav Kumar*, Tim Marks*, Wenxuan Mou*, Ye Wang, Michael Jones, Anoop Cherian, Toshiaki Koike-Akino, Xiaoming Liu and Chen Feng
CVPR 2020
pdf / supplementary / 1min talk / slides / demo / arxiv / code / dataset

Scaling Up Exact Neural Network Compression by ReLU Stability

Thiago Serra, Xin Yu, Abhinav Kumar and Srikumar Ramalingam
NeurIPS 2021
pdf / talk / code / openreview

Lossless Compression of Deep Neural Networks

Thiago Serra, Abhinav Kumar and Srikumar Ramalingam
CPAIOR 2020 and ICML LXAI Workshop 2020
pdf / slides / talk / poster / code

Zero Shot License Plate Re-Identification

Mayank Gupta*, Abhinav Kumar* and Sriganesh Madhvanath
WACV 2019 (Spotlight)
pdf / slides / poster / article

Fast and Generalisable License Plate Re-Identification using Neural Embedding of Fisher Vectors

Abhinav Kumar*, Shantanu Gupta*, Vladimir Kozitksy and Sriganesh Madhvanath
ICVGIP 2018 (Short Oral)
pdf / slides / arxiv

VPDS: An AI-based Automated Vehicle Occupancy and Violation Detection System

Abhinav Kumar*, Aishwarya Gupta*, Bishal Santra*, Lalitha Srinivasan*, Manasa Kolla*, Mayank Gupta* and Rishabh Singh*
IAAI (Application conference of AAAI) 2019
pdf / slides / article / tutorial / bronchure

Workshops

UGLLI Face Alignment: Estimating Uncertainty with Gaussian Log-Likelihood Loss

Abhinav Kumar*, Tim Marks*, Wenxuan Mou*, Chen Feng and Xiaoming Liu
ICCV Workshops on Statistical Deep Learning in Computer Vision 2019
pdf / slides / poster / media / Best Oral Presentation award

Others and Arxived

Equivalent and Approximate Transformations of Deep Neural Networks

Abhinav Kumar, Thiago Serra and Srikumar Ramalingam
pdf / talk / email for the code

Parametric Synthesis of Text on Stylized Backgrounds using PGGANs

Mayank Gupta, Abhinav Kumar and Sriganesh Madhvanath
pdf / demo

Estimation of Bandlimited Grayscale Images From the Single Bit Observations of Pixels Affected by Additive Gaussian Noise

Abhinav Kumar and Animesh Kumar
pdf / slides / report

Before 2016

SmartEvacTrack: A People Counting and Coarse-level Localisation Solution for Efficient Evacuation of Large Buildings

Nasim Ahmed, Avik Ghose, Amit Agarrwal, Chirabrata Bhaumik, Vivek Chandel and Abhinav Kumar
PERCOM Workshops on Crowd Assisted Sensing, Pervasive Systems and Communications 2015
link

Technology Enabled Assessment Environment

Rohit Kondekar, Pravin Kondekar, Deepak Phatak, Avinash Awate, Mayank Singh, Alefiya Lightwalla and Abhinav Kumar
IASTED International Conference on Technology for Education 2011
link / report

Patents

GrooMeD-NMS for Monocular 3D Object Detection

Xiaoming Liu, Abhinav Kumar and Garrick Brazil
US Patent filed 2021

Image Processing System and Method for Landmark Location Estimation with Uncertainty

Tim Marks, Abhinav Kumar, Wenxuan Mou, Chen Feng and Xiaoming Liu
US 16/592,822 A1, filed 2021
link

Estimation of Direction of Motion of Users on Mobile Devices

Avik Ghose, Abhinav Kumar, Chirabrata Bhaumik and Arpan Pal
US 10,520,330 B2, granted 2019
link

Talks

Mathematical Modeling for Landmark and 3D Object Detection

[link]
University of Utah