Abstract

Empathy plays a significant role in pervasive cooperation and prosocial behavior among heterogeneous groups of individuals in the animal kingdom. It modulates one's responses to the emotions and pain of others. It enables one to understand and share the feelings of others and then act accordingly and appropriately. The phenomenon of empathy has been researched in several areas such as philosophy, clinical, social and developmental psychology, social and neural sciences, etc. and still a mystery leading to interesting academic research. A plethora of factors have been found to modulate empathy. The aim of this study was to investigate how socioeconomic standing (in particular profession) of an individual (specifically women rape-victims) affects the empathic responses of the participant, both neuronally and behaviourally. This study focuses specifically on empathy towards pain as experienced by another. A three-part experimental study was conducted to explore the neural correlates of empathy in an individual when biased by the knowledge of the profession of the character (female rapevictim). The first part of this study focused on establishing if a bias existed and is reflected in behavioral responses in society towards different career professions opted by women. For this, we conducted an online survey with 21 professions (opted by women) and collected the respectability score. We observed a difference in the average respectability score for different professions thereby confirming the existence of a difference in attitudes towards different professions pursued by women. Progressing on the findings from the first part, the next stage focused on investigating the difference in empathic responses (behaviorally) of individuals towards the pain of others (female rape-victims). Towards this, a second survey was conducted where the participants were presented with the fictional incident of a rape incident narrated as post-incident trauma from the victim’s point of view. The narratives were designed to differ only in depictions of the profession of the victim. Standard questionnaires used to measure empathic responses were presented to participants in addition to the narrative specific questions. We found differences in empathic responses to perceived pain of victims as a function of the profession of the purported rape-victim.

Abstract

The significance of deep neural networks (DNNs) has been well established through its success in a variety of tasks. However, recent studies have shown that DNNs are vulnerable to adversarial examples — inputs crafted by adding small perturbations to the original input. Such perturbations are almost imperceptible to humans but deceive DNNs thus raising major concerns about their utility in real world applications. Although, existing adversarial attack methods in NLP have achieved high success rate in attacking DNNs, they either require detailed information about the target model, training data or need a large amount of queries to generate attacks. Therefore, such attack methods are not realistic as it does not reflect the types of attacks that can be encountered in the real world and are less effective as attacks relying on model information and excessive queries can be easily defended against. In this thesis, we address the above mentioned drawbacks by proposing two realistic attack settings — hard label black box setting and limited query setting. Next, we propose two novel attack methods that crafts plausible and semantically similar adversarial examples in the above settings. The first attack method uses a population based optimization procedure to craft adversarial examples in the hard label black box setting. The second method is a query efficient method that leverages word attention scores and locality sensitive hashing to find important words for substitution in the limited query setting. We benchmark our results across the same search space used in prior attacks so as to ensure fair and consistent comparison. To improve the quality of generated adversarial examples, we propose an alternative method that uses masked language model to find candidate words for substitution by considering the information of both the original word and its surrounding context. We demonstrate the efficacy of each of our proposed approach by attacking NLP models for text classification and natural language inference task. In addition to that we use adversarial examples to evaluate the robustness and generalization of recent math word problem solvers. Our results showcase that DNNs for the above tasks are not robust as they can be deceived by our proposed attack methods in a highly restricted setting. We conduct human evaluation studies to verify the validity and quality of generated adversarial examples.

Abstract

Walking in a virtual environment is a bounded activity. It is challenging to create and navigate in a large virtual environment given a limited physical space. Various techniques like walking on a treadmill, gesture-based walking and redirected walking address this problem in different ways. These existing solutions have challenges like dependency on additional hardware, inability to adopt different room sizes, compromised rendering quality etc. Such trade-offs make the VR setup immobile and expensive for VR applications which require endless locomotion of the participating subjects. We developed a new redirected walking method to allow the subject to walk continuously within a predefined play area without needing additional hardware support. It is a software-oriented system which dynamically generates an endless path for the participant to walk forward continuously. The system is adaptable to varied play area sizes and is highly mobile because no additional hardware setup is required. Redirected walking consists of methods which actively manipulate the virtual environment to fit it in a smaller physical space. Such techniques allow the user to explore a virtual environment larger than the physical space. Covert redirected walking and overt redirected walking are two sub-classes of the redirected walking techniques. Covert techniques manipulate the environment without being perceivable to the user, and overt redirection techniques do such manipulations while being noticed by the user. Our work is classified as an overt redirected walking technique. The system employs a path generation technique which is inspired by the path generation mechanics of mobile-based endless running games. Generally, the path is composed of successive chunks called “Tiles” in the endless running games. The path initially consists of a fixed number of tiles. When the player begins to move forward, a new tile is added towards the end and a tile is removed from the beginning of the path. This behaviour continues until the user terminates the game. Such an approach creates a perception of an infinitely long path and minimizes memory requirements. We designed an endless path generation system which dynamically generates and updates a path within the bounds of the play area. To evaluate the path generation system, we developed a virtual art gallery. The art gallery is an endless, dynamically generated corridor with paintings/art items hanging on the walls. The path generation system generates and updates the gallery when the user moves forward in the gallery. We conducted three user studies to validate our work and gather feedback to check whether the generated environment induces sickness in the subjects and whether it can adjust to different room sizes without affecting the user’s experience of presence and ease of locomotion. In the last and final user study, we employed a simulator sickness questionnaire (SSQ), a presence questionnaire (PQ) and a custom questionnaire (CQ). No significant differences were found in the scores of all three questionnaires between the two different room sizes. We found an overall low score in the experienced realism sub-scale of the PQ, indicating that the subjects didn’t perceive the generated environment as very real. CQ results suggest that the participants walked normally

Abstract

The world has been experiencing rapid urbanization over the last few decades, putting a strain on existing city infrastructure such as waste management, water supply management, public transport and electricity consumption. We are also seeing increasing pollution levels in cities threatening the environment, natural resources and health conditions. However, we must realize that the real growth lies in urbanization as it provides many opportunities to individuals for better employment, healthcare and better education. Cities have now become one of the major contributors to the world’s gross domestic product; therefore, urbanization should always be seen as an opportunity. However, it is imperative to limit the ill effects of rapid urbanization through integrated action plans to enable the development of growing cities. This gave rise to the concept of a smart city in which all available information associated with a city will be utilized systematically for better city management. Needless to say, data associated to the city plays a vital role in understanding the city’s current state and quickly addressing the issues discussed earlier. The systematic collection, processing and analysis of data allow for the incorporation of data intelligence and helps inform decisions on issues that matter to the lives of city residents. City management officials have been collecting data for years; however, their department-centric nature has created data silo issues and made it difficult to manage and share, thus hindering the data analysis process. Considering these issues in existing data systems, this work proposes a novel Smart City data management system architecture to systematically collect, store, exchange and analyze data collected from various sensing nodes depicting smart city applications. The proposed system architecture is divided in subsystems and is discussed in individual chapters. The first chapter introduces and gives overview to the reader of the complete system architecture. The second chapter discusses the data monitoring system (DMS) and data lake system (DLS) based on the oneM2M standards. DMS employs oneM2M as a middleware layer to achieve interoperability, and DLS uses a multi-tenant architecture with multiple logical databases, enabling efficient and reliable data management. The third chapter discusses energy monitoring and electric vehicle charging systems developed to illustrate the applicability of the oneM2M standards. The fourth chapter discusses the Data Exchange System (DES) based on the Indian Urban Data Exchange (IUDX) framework. DES uses IUDX’s standard data schema and open APIs to avoid data silos and enable secure data sharing. The fifth chapter discusses the 5D-IoT framework that provides uniform data quality assessment of sensor data with meaningful data descriptions.

Abstract

Making things, coming up with strange new ideas is what creativity is. The question of where these new ideas come from is the very question that drives all study of creativity. Creativity has to come from the knowledge that exists in the minds of the creator [1]. This work participates in the questions of how we come up with new ideas and to what extent the conceptual space influences creativity. We form concepts through the process of abstraction. Despite having been long established that creativity does not come from the Gods but from our mental and bodily processes, there is very little work on the processes of concept acquisition. There is almost no work that directly studies creativity and abstraction together. This work argues that creativity research assumes an Amodal Symbol cognition structure where the processes of cognition and higher cognitive functions are not related. As a result this relationship has not been explored. This work employs a mixed method approach, using both qualitative and quantitative methods for coding and analysis of abstractions and creative generations. A wordlist based on familiarity, level of abstractness and modality experience was created. A norming study of the rating of abstractness of these words showed significant variability in ratings indicating subjectivity in the perception of abstractness. To study abstractness and creativity a Property Generation and an Alternate Uses task was conducted. Traditional theories of creativity assume we manipulate taxonomic or entity properties in a computer like way, during the process of creativity, however our results emphasises the role of individual experiences in the process of creativity

Abstract

Machine Translation aided workflows have become increasingly prevalent across industry and within large institutional translation efforts. We evaluate the machine translation postediting process end-to-end on three aspects: the usefulness of the process itself, as to whether it leads to reduction in temporal, technical and cognitive effort compared to unaided translation; the linguistic nature of the products of this process, in terms of the quality of translations generated measured against unaided human translations; and the suitability of utilizing these outputs towards a standard machine translation evaluation task. The language direction we study is English-Hindi, a mid to high resource language pair. We first conduct a behavioral experiment utilizing professional translators. In our analysis, employing mixed-effects modeling techniques, we find that for this language pair, post-editing reduces translation time by as much as 63%, consequently increasing productivity; it reduces technical effort measured as keystrokes logged by 59%; and reduces cognitive burden measured as reduction in number of pauses by 63%. We then investigate the nature of the translations generated during post-editing by comparing them against raw MT outputs and unaided human translations on linguistic indicators we define and implement. We find significant differences between the three corpora in terms of lexical richness, normalization and interference by the source language. We also detect and confirm the presence of translation universals in our data. We then go on to test the suitability of translated data thus created for a machine translation evaluation task. We detect engine-reference bias towards the engine utilized during post-editing, inflated engine scores across the board on post-edited references, and find the utilization of multiple references to be the most prudent choice when conducting meta-evaluation tasks. We run these evaluations across multiple string-based and pre-trained metrics and issue general recommendations on constructing evaluation test sets and metrics to score against. We also contribute a set of post-editing guidelines with examples culled from a real-world translation task for translators and post-editors working with the English-Hindi pair.

Abstract

This thesis targets the problem of surrogate approximations for similarity measures to improve their performance in various applications. We have presented surrogate approximations for popular dynamic time warping (DTW) distance, canonical correlation analysis (CCA), Intersection-over-Union (IoU), PCP, and PCKh measures. For DTW and CCA, our surrogate approximations are based on their corresponding definitions. We presented a surrogate approximation using neural networks for IoU, PCP, and PCKh measures. First, we propose a linear approximation for the naïve DTW distance. We try to speed up the DTW distance computation by learning the optimal alignment from the training data. We propose a surrogate kernel approximation over CCA in our next contribution. It enables us to use CCA in the kernel framework, further improving its performance. In our final contribution, we propose a surrogate approximation technique using neural networks to learn a surrogate loss function over IoU, PCP, and PCKh measures. For IoU loss, we validated our method over semantic segmentation models. For PCP, and PCKh loss, we validated over human pose estimation models.

Abstract

Idiomatic expressions have always been a bottleneck for language comprehension and natural language understanding, specifically for tasks like Machine Translation(MT) and Natural Language Understanding(NLU). MT systems predominantly produce literal translations of idiomatic expressions as they do not exhibit generic and linguistically deterministic patterns which can be utilized for the comprehension of the non-compositional meaning of the expressions. These expressions occur in parallel corpora used for training, but due to the comparatively high occurrences of the constituent words of idiomatic expressions in a literal context, the idiomatic meaning gets overpowered by the compositional meaning of the expression. The absence of data with a large coverage and quantity of idiomatic expressions exacerbates the issue to handling them further. Our work aims to provide a method of handling idiomatic expressions which not only suggests a pipeline for the task but also enables a process of data creation from subsequent steps which can be used in further downstream tasks. State of the art metaphor detection systems are able to detect non-compositional usage at word level but miss out on idiosyncratic phrasal idiomatic expressions. This creates a dire need for a dataset with a wider coverage and higher occurrence of commonly occurring idiomatic expressions, the spans of which can be used for Metaphor Detection. With this in mind, we present our English Possible Idiomatic Expressions(EPIE) corpus containing 25206 sentences labelled with lexical instances of 717 idiomatic expressions. These spans also cover literal usages for the given set of idiomatic expressions. We also present the utility of our dataset by using it to train a sequence labelling module and testing on three independent datasets with high accuracy, precision and recall scores. Natural Language Understanding has made recent advancements where context-aware token representation and word disambiguation have become possible to a large extent. In this scenario, comprehension of phrasal semantics particularly in the context of multi word expressions (MWE) and idioms, is the subsequent task to be addressed. Word level metaphor detection is unable to handle phrases or MWE(s) which occur in both literal and idiomatic context. State of the art transformer architectures can be useful in this context, but the absence of a large comprehensive dataset is a bottleneck. In this paper, we present a labelled EPIE dataset containing 3136 occurrences for 358 formal idioms. To prove the efficacy of our dataset, we also train a sequence classification model effectively and perform crossdataset evaluation on three independent datasets. Our method achieves good results on all datasets with F1 score of 96% on our test data, and 82%, 74% and 76% F1 score on SemEval All Words, SemEval Lex Sample, and PIE Corpus datasets respectively

Abstract

With the booming online market, managing the warehouse inventory is one of the most essential and challenging tasks. The management of inventory will be more efficient if it is automated using robots. The robots can work faster than humans, the robots work at a constant speed with no breaks, and do tasks in more repetition than humans like fetching inventory from warehouse. But for the robots to perform tasks like putting objects in racks, fetching objects from rack, re-organising the racks to make more space they need to have an understanding of the warehouse environment. To understand the warehouse environment the robots needs to create a 3D map of the warehouse consisting space to move for the robot as well as identify the racks and boxes so that its able to plan the execution of tasks. Towards solving this problem in this thesis we address problem of freespace estimation for rack shelves. Given a monocular RGB image captured from a camera mounted on a robotic arm. We aim to predict the Top-view and Front-view layouts so as to create a 3D reconstruction of rack and objects present in the Monocular RGB image. We propose a simple yet effective network architecture RackLay, which takes a monocular RGB image as input and outputs the Top-view and Front-view layout of all the shelves comprising the rack visible in the image. The Network can learn two kinds of layout representations, one in the canoncial frame centered on the shelf, called the shelf-centeric layout and the other in a frame with respect to the camera, called the ego-centric layout. Apart from portraying the versatility of the network, they lend to various useful applications. Since there are very few publicly available datasets for warehouse settings, we also introduce the synthetic data generation pipeline termed as WareSynth, which can be used to generate 3D warehouse scenes, automate the process of data capture and generate corresponding annotations. WareSynth can be used for various tasks such as 2D/3D object detection, semantic/instance segmentation, layout estimation, 3D scene navigation and mapping, 3D reconstruction etc. The same pipeline can also be modified to other kind of scenes such as supermarkets, greenhouses by changing the database of objects and placement parameters hence this pipeline open gates for further research in similar environments.

Abstract

Blockchains lie at the heart of Bitcoin and other cryptocurrencies that have shown great promise to revolutionize finance and commerce. The novel applications of blockchain technology are derived from the unique combination of properties of blockchains: immutability, incorruptibility, consensus, transparency, decentralization and ordering. However, they face technical challenges when it comes to scaling to support greater demand while maintaining their desirable security properties. In an exciting line of recent work, many researchers have proposed various scalable blockchain protocols that demonstrate the potential to solve these challenges. However, many of these protocols come with the assumptions of honest majority and symmetric network access which may not accurately reflect the real world where the participants may be self-interested or rational. In this thesis, we work towards analysing blockchain protocols in settings with rational participants and use game-theory to predict their behaviour. For the first time in literature, we study blockchain protocols in a setting with asymmetric network access and highlight the role of network fairness in ensuring the security of a blockchain protocol. We demonstrate via simulations, the effect of lack of network fairness in Bitcoin and the equilibrium behaviour of the participants which results in loss of both security and performance of the protocol. In order to further highlight the importance of network fairness in designing scalable blockchain protocols, we showcase incentive-driven deviations in the OHIE blockchain protocol proposed recently and the loss of security caused by these deviations. We then study a class of blockchain protocols designed by layering two distinct types of consensus protocols and determine the parameters required to ensure game-theoretic soundness and security. We also formally study the reward schemes employed by Proof-of-Work-based blockchain protocols and showcase the lack of game-theoretic soundness. Towards this, we also propose a class of game-theoretically sound reward schemes. Finally, we extend our work by developing a model for layered-blockchain protocols and present a game-theoretic model for analyzing the security of layered-blockchain protocols. Therefore, in this thesis we present novel work that aims to guide the development of fair and game-theoretically sound blockchain protocols.