Accepted Papers

Session 1

#169. Ariel Kwiatkowski, “Creating Interactive Crowds with Reinforcement Learning”

  • Abstract: The entertainment industry, as well as the field of Computer Graphics (CG), frequently faces the issue of creating large crowds of people that would populate a certain scene. One of the ways to achieve that, particularly with modern rendering technique, is by using simulation – this, however, is nontrivial to design and control by the artists. The main goal of my PhD work is working towards the creation of a tool enabling the creation of virtual crowds that one can interact with, and we believe the best way to that is through Multiagent Reinforcement Learning (MARL) techniques. These animated crowds can then be used both in movies and video games. Especially for the latter, it is highly desirable that both the crowd as a whole, as well as the individual characters, can react to the user’s input in real time.

#180. Lydia Bryan-Smith, “Using Multimodal Data and AI to Dynamically Map Flood Risk”

  • Abstract: Classical measurements and modelling that underpin present flood warning and alert systems are based on fixed and spatially restricted static sensor networks. Computationally expensive physics-based simulations are often used that can’t react in real-time to changes in environmental conditions. We want to explore contemporary artificial intelligence (AI) for predicting flood risk in real time by using a diverse range of data sources.

#436. Sebastian Berns, “Increasing the Diversity of Deep Generative Models”

  • Abstract: Generative models are used in a variety of applications that aim for the production of diverse output. Currently, models are optimised for sample fidelity and mode coverage. My future work aims to increase their output diversity. In previous work, I have analysed the limitations of generative models and studied their use in artistic settings, as well artistic strategies to overcome their limitations.

#305. Nicholas Halliwell, “Evaluating Explanations of Relational Graph Convolutional Network Link Predictions on Knowledge Graphs”

  • Abstract: Recently, explanation methods have been proposed to eval- uate the predictions of Graph Neural Networks on the task of link prediction. Evaluating explanation quality is difficult without ground truth explanations. This thesis is focused on providing a method, including datasets and scoring metrics, to quantitatively evaluate explanation methods on link pre- diction on Knowledge Graphs.

Session 2

#136. Oshin Agarwal, “Towards Robust Named Entity Recognition via Temporal Domain Adaptation and Entity Context Understanding”

  • Abstract: Named Entity Recognition models perform well on benchmark datasets but fail to generalize well even in the same domain. The goal of my thesis is to first quantify the generalization capabilities of NER models by establishing appropriate benchmarks, metrics and best practices for task setup. I also aim to understand the reason for such lack of robustness by probing whether models memorize entity names or they can recognize predictive contexts. Finally, with the developed resources and the insights gained from the analysis, I seek to improve the robustness of NER models, focusing on the recognition of ethnically diverse entities and new entities over time when these mod- els are deployed in practice. My work will answer research questions pertaining to the robustness of these models. It will also have practical implications, particularly the techniques for improving robustness to evolving entities and language over time, can be adopted by NER practitioners.

#101. Kanishka Misra, “On Semantic Cognition, Inductive Generalization, and Language Models”

  • Abstract: My doctoral research focuses on understanding semantic knowledge in neural network models trained solely to predict natural language (referred to as language models, or LMs), by drawing on insights from the study of concepts and categories grounded in cognitive science. I propose a framework inspired by `inductive reasoning,’ a phenomenon that sheds light on how humans utilize background knowledge to make inductive leaps and generalize from new pieces of information about concepts and their properties. Drawing from experiments that study inductive reasoning, I propose to analyze semantic inductive generalization in LMs using phenomena observed in human-induction literature, investigate inductive behavior on tasks such as implicit reasoning and emergent feature recognition, and analyze and relate induction dynamics to the learned conceptual representation space.

#344. Rohan Paleja, “Mutual Understanding in Human-Machine Teaming”

  • Abstract: The aim of my thesis is to (1) enable collaborative robots (i.e., cobots) to understand heterogeneous end-users using white-box learning (xAI) methods, (2) identify the contexts in which such xAI methods can help improve HMT, and (3) enable cobots to efficiently communicate information that supports HMT. By tackling several key challenges within HMT, we augment cobots with the ability to more effectively team with human partners.

#175. Indrapriyadarsini Sendilkkumaar, “On the practical robustness of the Nesterov’s accelerated quasi-Newton method”

  • Abstract: This study focuses on the Nesterov’s accelerated quasi-Newton (NAQ) method in the context of deep neural networks (DNN) and its applications. The thesis objective is to confirm the robustness and efficiency of Nesterov’s acceleration to quasi-Netwon (QN) methods by developing practical algorithms for different fields of optimization problems.

Session 3

#397. Deeksha Arya, “AI-Driven Road Condition Monitoring Across Multiple Nations”

  • Abstract: The doctoral work summarized here is an application of Artificial Intelligence (AI) for social good. The successful implementation would contribute towards low-cost, faster monitoring of road conditions across different nations, resulting in safer roads for everyone. Additionally, the study provides recommendations for re-using the road image data and the Deep Learning models released by any country for detecting road damage in other countries.

#277. Shefeh Mbuy, “Dynamic Algorithmic Impact Assessment to Promote an Ethical Use of AI in Businesses”

  • Abstract: My PhD research focus is to produce a critical review of literature in Algorithmic Impact Assessment (AIA) and to develop an AIA tool that can be used to evaluate potential unintended impact of AI systems.

#319. Denizalp Goktas, “An Algorithmic Theory of Markets and their Application to Decentralized Markets”

  • Abstract: Broadly speaking, I hope to dedicate my PhD to improving our understanding of algorithmic economics with the ultimate goal of building welfare-improving technology to create more efficient markets. In this application, I describe how my past work has built on the existing literature to get closer to the goal of creating such technologies, and describe what research paths this work opens up for the rest of my PhD. I believe that my research has the potential to provide algorithmic solutions to problems in machine learning, optimization, and algorithmic game theory, and can be used to improve the efficiency of online marketplaces.

#434. Stefan Heidekrueger, “Equilibrium Learning in Auction Markets”

  • Abstract: My doctoral thesis aims to understand the theory and practice of learning Bayesian Nash equilibria in auctions. Positive results would open the door to wide-ranging applications in Market Design and the economic sciences.

Session 4

#104. Aarti Malhotra, “Socially Intelligent Affective AI”

  • Abstract: Artificial Intelligence has aimed to give artificial systems or agents, the ability to learn, perceive, recognize, plan, reason and act. Affective Computing has brought into focus the importance of giving AI systems, the capability to perceive, detect, utilize and generate emotion, affect, sentiment or feelings. To have a meaningful and rich human-computer interaction, we need to make AI more socially intelligent and affective. The doctoral research goal is to attempt to touch upon some of these aspects, firstly by surveying computational models implemented in AI that uses emotion in decisionmaking or behaviour; secondly, by creating new model to predict social event and affect in group videos; thirdly, to predict the social identities in visual scenes,; and lastly to predict social incoherence and to recommend appropriate behaviour.

#205. Matthew Fontaine, “Towards Automating the Generation of Human-Robot Interaction Scenarios”

  • Abstract: My thesis work studies the problem of generating scenarios to evaluate interaction between humans and robots. I expect these interactions to grow in complexity as robots become more intelligent and enter our daily lives. However, the limitations of evaluating such interactions \emph{only} through user studies, which are the de facto evaluation method in human-robot interaction, will quickly become infeasible. Therefore, I propose automatically generating scenarios to explore the diverse possibility space of scenarios to better understand interaction and avoid costly failures in real world settings.

#324. Fan Meng, “Creating Interpretable Data-Driven Approaches for Tropical Cyclones Forecasting”

  • Abstract: Tropical cyclones (TC) are extreme weather phenomena that bring heavy disasters to humans. Existing forecasting techniques contain computationally intensive dynamical models and statistical methods with complex inputs, both of which have bottlenecks in intensity forecasting, and we aim to create data-driven methods to break this forecasting bottleneck. The research goal of my PhD topic is to introduce novel methods to provide accurate and trustworthy forecasting of TC by developing interpretable machine learning models to analyze the characteristics of TC from multiple sources of data such as satellite remote sensing and observations.

#190. Geetanjali Bihani, “Interpretable Privacy Preservation of Text Representations Using Vector Steganography”

  • Abstract: Contextual word representations generated by language models learn spurious associations present in the training corpora. Adversaries can exploit these associations to reverse-engineer the private attributes of entities mentioned in the training corpora. These findings have led to efforts towards minimizing the privacy risks of language models. However, existing approaches lack interpretability, compromise on data utility and fail to provide privacy guarantees. Thus, the goal of my doctoral research is to develop interpretable approaches towards privacy preservation of text representations that maximize data utility retention and guarantee privacy. To this end, I aim to study and develop methods to incorporate steganographic modifications within the vector geometry to obfuscate underlying spurious associations and retain the distributional semantic properties learnt during training.