Anoop Krishnan

Visual Computing and Biometric Security Lab (VCBSL) Wichita State University. 1845, Fairmount St.Wichita . Kansas . 67260 axupendrannair@shockers.wichita.edu

I am a PostDoctoral Researcher at Cox Business School, Southern Methodist University. I do research on Fair and Ethical Biometric Systems, which involves studying about the biases in the present systems and mitigate it through deep learning techniques. Lab Webpage.

Publications

Understanding Fairness of Gender Classification Algorithms across Gender-Race groups

Authors:Anoop Krishnan,Ali Almadan, Ajita Rattani
19th ICMLA 2020 | Miami, Florida

BWCFace: Open-set Face Recognition using Body-worn Camera

Authors:Ali Almadan,Anoop Krishnan, Ajita Rattani
19th ICMLA 2020 | Miami, Florida

Probing Fairness of Mobile Ocular Biometrics Methods Across Gender on VISOB 2.0 Dataset

Authors:Anoop Krishnan*,Ali Almadan*, Ajita Rattani
25th ICPR 2020| Milan, Italy

A Survey of Biometric and Machine Learning Methods for Tracking Students’ Attention and Engagement

Authors:Maria Villa, Mikhail Gofman, Sinjini Mitra,Ali Almadan,Anoop Krishnan, Ajita Rattani
19th ICMLA 2020 | Miami, Florida

Investigating Fairness of Ocular Biometrics Among Young, Middle-Aged, and Older Adults

Authors:Anoop Krishnan,Ali Almadan, Ajita Rattani
ICCST 2021, University of Hertfordshire, Hatfield, UK

Experience

Graduate Research Assistant

advised by Dr. Ajita Rattani, Department of EECS, Wichita State University
Research is being done on understanding the biases on different gender classification models. All models are implemented using Pytorch. For more information, refer the publication.

Aug 2020 - present

Reviewer

ACM Multimedia Conference 2020

12-16 Oct 2020

Reviewer

IET Biometrics

2021

Graduate Teaching Assistant

Department of EECS, Wichita State University
Assisted graduate students at Image Analysis and Computer Vision practice(Python) and undergraduate students in Introduction to Software Engineering course(As a Grader)

Jan 2020 - May 2020

Teaching Assistant

Department of Computer Science and Engineering, Indian Institute of Information Technology Design & Manufacturing, Kancheepuram

Assisted undergraduate students at VLSI lab, Object Oriented Programming Practice(C++) and Digital Image processing practice(Matlab)

July 2018 - April 2019

Project Intern

FISST Pvt Ltd, Chennai, India

evelopment of an automatic number plate recognitionsystem for toll plaza environment and considering only Indian vehicles. A Convolutional Neural Network based model for plate detection and a CRNN based model for character recognition had modeled and interfaced through Flask. Python was used forimplementation. Colab has been used for training.

May 2018 - April 2019

Intern

Tata Consultancy Services, Chennai, India

Design of digital application for inculcating design thinking in organizations.

Dec 2015 - July 2016

Education

Wichita State University

Department of Electrical Engineering and Computer Science
Doctor of Philosophy (Ph.D.) - EECS

Jan 2020 - present

Indian Institute of Information Technology Design and Manufacturing, Kancheepuram

Department of Computer Science & Engineering
Dual Degree - Computer Engineering

July 2014 - July 2019

Oxford Senior Secondary School, Karavaloor

All India Senior School Certificate Examination

May 2011 - May 2013

Blogs

Where do Models go Wrong? Parameter-Space Saliency Maps for Explainability

Roman Levin, Manli Shu, Eitan Borgnia, Furong Huang, Micah Goldblum, Tom Goldstein

The authors try to understand the network parameters, and come up with a parameter saliency map. This map reveals the network component which malfunctions, and thus pruning it, would improve the model behavior.The proposed method to map the parameter saliency is to use the gradient information of the loss function as a measure of parameter sensitivity and optimality of thr measure at a give point in image space. The parameter saliency profile allows us to identify filters that are most responsible for mistakes and erroneous network behavior. The authors create a framework for identifying the particular filters that cause incorrect predictions and examining the relationships between these filters and images. This approach results in a comprehensible and intuitive knowledge of model behavior.

Face Quality Estimation and Its Correlation to Demographic and Non-Demographic Bias in Face Recognition

Title description, Sep 2, 2017

Interests

Apart from being a researcher, I enjoy playing badminton and I do passionate about photography. Also spend my time with my friends.

When forced indoors, I follow a number of sci-fi, feel-good genre and highly appreciated movies and television shows, I do love to cook, and I spend a large amount of my free time exploring the latest advancements in biometric systems.

Awards & Certifications