Vishaal Udandarao

I am a first year ELLIS PhD student, jointly working with Matthias Bethge at The University of Tuebingen and Samuel Albanie at The University of Cambridge. I am mainly interested in research problems that involve understanding, interpreting and representing (un)structured data using machine (deep) learning. I am especially passionate about adapting large-scale vision-language models in a continual learning fashion. Previously, I was an MPhil Machine Learning and Machine Intelligence student at The University of Cambridge. I graduated from IIIT Delhi with a B.Tech in Computer Science & Engineering in July, 2020.

At IIIT Delhi, I worked with Dr Saket Anand in the Infosys Center for Artificial Intelligence (CAI). My work revolved around disentangling latent spaces of variational auto-encoders, in unsupervised and weakly supervised settings to capture robust and domain-invariant latent features of images efficiently. I also interned with the Rutgers Machine Learning Lab (RUML) advised by Dr Sungjin Ahn where I worked on object-centric representation learning and unsupervised continual learning. I am also fortunate to have worked with Dr Rajiv Ratn Shah and Rajesh Kumar on privacy leakage problems and their correlations with behavioural biometrics from keystroke and swiping data. I completed my undergraduate thesis under the joint guidance of Dr Anubha Gupta and Dr Tanmoy Chakraborty at the Signal Processing and Biomedical Imaging Lab (SBILab). My thesis involved applications of AI for education. Prior to joining IIIT Delhi, I completed my high school education from Indian School Al Wadi Al Kabir, Muscat.

Apart from my academic interests, I am a huge football fan and actively support FC Barcelona. Fun fact: Did you know Lionel Messi is the greatest player to ever touch a football? No? Well, deal with it. I also love watching Formula 1 and look up to Lewis Hamilton. I used to write stuff, but that was a long long time ago. I also dabble around with the guitar and the keyboard at times. Checkout my soundcloud profile!

Email  /  CV  /  Google Scholar  /  Twitter  /  Github

A VAE model that learns implicit structure by aligning the latent space with the Valence-Arousal circumplex space. Further, a novel algorithm for mapping categorical and dimensional model labels using annotation transfer across affective facial image datasets is depicted.

A novel bi-modal framework that aims to train two modalities (image and text) in a joint fashion inspired by the Contrastive Predictive Coding (CPC) and Noise Contrastive Estimation (NCE) paradigms which preserve both inter and intra-class relationships in a modality-invariant fashion.

A multi-task learning model which incorporates auxiliary semantics by utilising a weight alignment layer and information exchange layer.

A self-supervised framework to disentangle multiple attributes by exploiting structural inductive biases within images and leveraging contrastive learning paradigms.

An empirical study on the inference of gender, major/minor (computer science, non-computer science), typing style, age, and height from the typing patterns collected from 117 individuals in a multi-device environment.

A meme generation system that uses a trained state-of-the-art transformer-based (GPT-2) model for caption generation by employing an encoder-decoder architecture.

An on-the-fly conceptual network model that incorporates educational semantics and preserves correlations between conceptual entities by applying intelligent indexing algorithms on an inherent concept network so as to improve answer generation.