Talks and Presentations

Private-Share: A Secure and Privacy-Preserving DeCentralized Framework for Large Scale Data Sharing

December 01, 2021

Talk, ACM MM Asia 2021 (Virtual), Australia

The various data and privacy regulations introduced around the globe, require data to be stored in a secure and privacy-preserving fashion. Non-compliance with these regulations come with major consequences. This has led to the formation of huge data silos within organizations leading to difficult data analysis along with an increased risk of a data breach. Isolating data also prevents collaborative research. In this presentation we show how we address this through Private-Share, a framework that would enable secure sharing of large scale data. [Presentation Recording]

Deepfakes for Histopathology Images: A Myth or Reality?

April 30, 2021

Talk, Mizzou Cybersecurity Center IAB Meeting (Virtual), Columbia, Missouri

Deepfakes have become a major public concern on the Internet as fake images and videos could be used to spread misleading information about a person or an organization. The field of digital pathology is gaining a lot of momentum as the Food and Drug Administration (FDA) is approving new digital pathology systems for primary diagnosis and consultation in the US. Hence, in this presentation, we explore if deepfakes can be generated for histopathology images using advances in generative adversarial networks (GANs) and deep learning in general.

A Large-Scale Image Retrieval System for Everyday Scenes

March 09, 2021

Talk, ACM MM Asia 2020 (Virtual), Singapore

In this presentation, we showcase a large-scale image retrieval system for everyday scenes. The system leverages advances in deep learning and natural language processing (NLP) for image understanding tasks. The system allows a user to retrieve highly relevant images by leveraging relationships between the objects within an image. We also demonstrate how the system can let a user to intuitively explore the repository and obtain suggestions for similar image queries to further explore the repository. [Presentation Recording]

QIK: A System for Large-Scale Image Retrieval on Everyday Scenes With Common Objects

October 29, 2020

Talk, ACM ICMR 2020 (Virtual), Dublin, Ireland

In this presentation, we propose a system for large-scale image retrieval on everyday scenes with common objects by leveraging advances in deep learning and natural language processing (NLP). Unlike recent state-of-the-art approaches that extract image features from a convolutional neural network (CNN), our system exploits the predictions made by deep neural networks for image understanding tasks. Our system aims to capture the relationships between objects in an everyday scene rather than just the individual objects in the scene. We also present the performance of our system on the Microsoft COCO dataset containing everyday scenes (with common objects) and prove that our system can outperform state-of-the-art techniques in terms of mean average precision for large-scale image retrieval. [Presentation Recording]

Deep Learning for Energy Usage Prediction in Sustainable Building Design

October 29, 2020

Talk, NSF CBL Semiannual IAB Meeting (Virtual), Kansas City, Missorui

As per the 2030 challenge, all new buildings, developments, and major renovations shall be carbon-neutral by 2030. In this talk, as a step forward towards achieving sustainable design of buildings we present our objective of developing a unified and accurate deep learning-based model for predicting energy usage of different building types. In addition to it, we also present our evaluation results compared against traditional machine learning techniques.

QIK: Querying Images via Knowledge Graphs

October 17, 2019

Talk, NSF CBL Semiannual IAB Meeting, Gainesville, Florida

The global image recognition market is expected to reach $40 billion by 2021. There is a growing demand for intelligent image retrieval solutions for specialized domains. With this in mind, we design QIK, a system for large-scale image retrieval. As an update to the Spring 2019 Semiannual meeting, in this presentation, we describe the outcomes in terms of the system evaluation and a few search results from QIK.

Arun Zachariah