Abstract

This thesis addresses the critical challenge of improving road safety by introducing novel approaches to predictive modeling of accident-prone zones and action recognition in critical traffic scenarios. It makes two key contributions: the early identification of accident-prone zones using Advance Driving Assistance System (ADAS) data and the development of IDD-CRS, a comprehensive dataset for action recognition in unstructured road environments. In the first study, geo-tagged collision alert data from a fleet of 200 ADAS-equipped city buses in Nagpur, India, is leveraged to proactively identify high-risk zones across urban road networks. Using Kernel Density Estimation (KDE), this study captures the spatiotemporal distribution of collision alerts, enabling the detection of emerging blackspots before accidents occur. A novel recall-based metric evaluates the alignment of these predicted zones with historical blackspots, while Earth Mover Distance (EMD)-based analysis identifies previously unreported accident-prone areas. This predictive framework provides civic authorities with actionable insights for targeted interventions, such as traffic-calming measures and infrastructure improvements, thereby enhancing public safety. The second part of the thesis introduces the IDD-CRS dataset, a large-scale collection of traffic scenarios recorded using ADAS and dash cameras. IDD-CRS fills a critical gap in existing datasets by focusing on complex interactions between vehicles and pedestrians, with scenarios such as high-speed lane changes, unsafe vehicle approaches, and near-miss incidents. With precise temporal annotations powered by ADAS technology, the dataset ensures accurate event boundaries, providing a robust benchmark for action recognition and long-tail action recognition tasks. It includes 90 hours of footage spanning 5,400 one-minute videos and 135,000 frames, with hard negative examples to challenge existing models. Initial benchmarks highlight the limitations of current video backbones in recognizing rare events, emphasizing the need for further advancements. Together, these contributions provide a holistic framework for improving road safety through proactive accident prevention and robust action recognition in traffic scenarios. By addressing both spatial accident prediction and temporal event recognition, this work offers foundational resources and actionable insights to advance research and practical solutions for safer road environments.

Abstract

Understanding what makes certain stimuli memorable is a critical area of research that has implications in two major industries: advertisements and education. The ability to create memorable essays and other textual materials on different topics are of great importance in education. Similarly, in the entertainment industry, there is a need to identify short video clips that are memorable to content consumers. This thesis investigates the concept of memorability in two distinct modalities: words and videos. The general topic under study is the factors influencing memorability, with a specific focus on how features such as semantic network properties and word properties affect word memorability, and how spatio-temporal attention mechanisms influence video memorability. The central questions addressed are: What makes certain words and videos more memorable than others? How do cognitive and cultural factors impact memorability across different languages? Previous research has explored various aspects of word and video memorability, but there is a gap in understanding the cross-linguistic differences in word memorability and the underlying human vs. model mechanisms for video memorability. In our first exploration, we conduct a cross-linguistic study comparing the memorability of words in English and Hindi. Our analysis reveals that centrality metrics in semantic networks are weakly indicative of word memorability in English, as well as in Hindi. Moreover, the memorability of words in Hindi and English were found to be uncorrelated, suggesting that language-specific cognitive and cultural factors might play a significant role in memorability. Word frequency remains a strong factor influencing memorability in both English and Hindi. This research highlights the need to consider both intrinsic and extrinsic effects when studying memorability. In the second exploration, we leveraged a simple CNN+Transformer model to predict video memorability. This model not only achieved state-of-the-art performance but also provided insights into spatio-temporal attention patterns. Human gaze data was collected to compare with the model’s attention mechanisms, showing that both humans and the model focus on similar regions of the videos, both spatially and temporally. This finding mirrors human attention patterns and provides a deeper understanding of the temporal dynamics in video memorability. Additionally, our analysis identified critical issues in existing video memorability datasets, offering guidelines for future dataset construction and validation.The significance of this research lies in its contribution to the understanding of memorability across different modalities and languages. Future research can build on these findings to enhance the robustness and generalizability of memorability models, as well as build foundations for endto-end products. A few examples include helping students learn better by making memorable educational content, big companies understanding their consumer base better by analysing advertisement content, and so on

Abstract

This work focuses on applying task-specific deep learning models to the brain encoding for both language and speech processing. Brain encoding involves developing models that predict the response of the brain to stimuli, which could reveal some understanding of how the brain processes the information presented. Previous work showed it was possible to predict brain responses successfully using selfsupervised language and speech models like Bert, Wav2Vec2.0, etc., but little examination has been conducted to determine whether further improvement can be achieved by using fine-tuned task-specific models. In this work, we rely on a variety of transformer-based models fine-tuned for different speech and language tasks—such as automatic speech recognition (ASR), phoneme recognition (PR), and sentiment analysis (SA)—to make predictions on neural activity. Such task-specific models should better capture the fine details of brain activity during listening to stories or reading sentences. Our method trains encoding models using these fine-tuned representations to predict responses in fMRI for different regions of the brain, including the whole brain, as well as specific language and auditory regions. This work introduces an overarching framework that systematically benchmarks the performance of task-specific representations in predicting brain activity. This is then demonstrated with two very different datasets—one from the reading of sentences (Pereira dataset) and the other from listening to spoken stories (Narratives dataset). For both datasets, we observe that the best performance in terms of predicting whole brain activity as well as language regions corresponds to models fine-tuned on ASR for speech encoding and Named Entity Recognition (NER) for language encoding. Further differences were observed in task representations that were maximally associated with brain responses related to reading versus listening. We also explored cross-modal functioning by comparing how the brain responds differently when processing linguistic stimuli through reading versus listening. The key contributions of this work are brain encoding taskonomy, which categorizes tasks along the lines of how well they align with neural responses. We analyze abstract versus concrete concepts and bring out different neural pathways for the same. Our results demonstrate the unique contributions of semantic tasks such as paraphrase generation and summarization and syntactic tasks such as coreference resolution and named entity recognition towards brain encoding while providing insight into neural correlates for language comprehension and auditory processing. This research advances cognitive neuroscience by illuminating how representations, especially taskspecific representations derived from deep learning models, can further elucidate encoding mechanisms in the brain. It bridges AI and neuroscience; potential applications could actually mean a breakthrough in brain-computer interfaces and cognitive technologies

Abstract

Current pedagogical systems in dance majorly focus on performance training which has a major leaning towards entertainment service. As a result, the repertoire of movements has been restricted in catering to what is popular. This has led to a repetitive template-based training fitting into the audio-visual norms of the aesthetics of the time, which restricts the scope for further evolution of the dance form and curtails the possibilities of newer dance forms to emerge. Historically dance as a knowledge system was fundamentally taught in a form-neutral, styleneutral fashion with the ulterior motive of serving broad societal functions including ritualistic, devotional, and entertainment purposes. It is important to observe that only with the universal understanding of dance, the masters of yesteryears were able to design and formulate newer artforms in order to meet newer requirements of the clients and times. Any knowledge system has its universal fundamentals through which any of its regional practices can be taught. For example, music is generally taught through musical notes, considered to be universal and allows any musical system to be taught and practised. However, dance does not have common fundamentals through which all forms of dances can be taught. This research aims at deriving these universal fundamentals of dance systematically and devise a pedagogical system to disseminate these universals. To this end, the objectives are as follows. First, to establish the possibility of having universal dance fundamentals by reviewing the pertinent Dance Texts and understanding Dance Practices of Traditional Practitioners. Second, to systematically arrive upon categories of universal dance fundamentals to create form-neutral, style-neutral holistic Indian Dance Pedagogy catering to children and beginners. Third, to create a Pedagogical method to teach the universal dance fundamentals. Fourth, to develop a pedagogical tool as an aid to teach the universal dance fundamentals. Since the lower half body training serves as the fundamental training offered in any dance form, this research is focussed in creating a pedagogical system to understand the dance movements of the lower half of the body. In order to identify the universal fundamentals of dance, a survey of the lower-half bodypostures and movements mentioned in the Ancient Indian Dance texts such as Nāṭyaśāstra, Saṅgītaratnākara and Abhinaya Darpaṇa, and specific postures where the foot/feet are in contact with the floor were collected. As the major dance texts are not form- or style-specific, the postures mentioned in the dance texts are form-neutral and style-neutral. Based on analysis of these texts and from practice, the logical progression from one posture to another is identified. This gives rise to the logical prediction of the unmentioned postures in between the already mentioned postures of the texts. This enables a learner not to memorise the postures but gives a pedagogical method to understand movements as a progression based on logical parameters. Similarly the movements from the dance texts and practices were collected and logically arranged. In the process, it is observed that movements have three major components, that is, Posture, Transition and Rhythm in Transition. Posture is the static body position which serves as a marker to the movements. This way of understanding Postures springs out of an anatomical approach and the possible affordances of the joints, to make a posture. The postures are systematically arrived at based on eight parameters namely, part of the body in contact with the earth, Knee Positions, Knee angle, Feet angle, Body weight orientation, Foot position, Part of the foot touching the point and Feet Distance. Transition is dynamic. It is understood as a transitory movement from one posture to another. Despite similar postures across sports, martial arts, yogasana, dance and theatre could be common, the Transition, or the movement in between the postures is the factor which decides whether the act is of sports, martial arts, yogasana, dance or theatre. It is in fact the Transition which decides the form or style of dance. The Transitions are systematically arrived at based on seven parameters namely, lifted leg, point of lifting the leg, foot variation which is touching the point of lifting, part of the foot which is touching the point, trajectory traced by the foot as it reaches to the point, trajectory traced by the foot as it comes out of the point of lifting and foot activity. Furthermore, the parameter Rhythm in Transition determines the specific dance form and dance style. By understanding Postures, Transitions and Rhythm in Transition one can systematically understand the fundamentals of the movements, and more specifically dance movements. The mentioned three parameters served as a basis for a pedagogical framework, based on which a pedagogical meth

Abstract

This research work presents the design and FPGA implementation of a comprehensive radar signal processing system, focusing on de-interleaving, tracking, and pulse parameter estimation of multiple radar signals. The system encompasses end-to-end signal processing, starting from the down-converted intermediate radar signal as input, which is then processed to generate pulse descriptive words (PDW) containing crucial information such as frequency, bandwidth, pulse width, 3 Dimensional Direction of arrival (DOA), and pulse repetition interval (PRI), including PRI encoding technique. The system is capable of de-interleaving the signals of distinct types such as Pulse radar signals, frequency modulated continuous waves. The pulsed radar signals can be encoded in any type of PRI scheme such as fixed, jitter, and staggered PRI encoding techniques. Notably, the system is designed to track up to six non-coherent radar sources concurrently, providing situational awareness and signal detection capabilities. The accuracy achieved in the direction of arrival estimation is remarkable, with an error margin of less than 1°, for the signals with a signal-to-noise ratio (SNR) greater than 20dB. The experimental results obtained from the FPGA implementation demonstrate the effectiveness of the proposed system in comparison to other DOA estimation algorithms implemented on Field Programmable Gate Array (FPGA). The FPGA-based implementation of the De-interleaving of radar signals is tested in an anechoic chamber. The entire radar processing system is implemented using Verilog Hardware descriptive language. The test signal is obtained from the radar signal generator and the verification, validation is performed on Xilinx Virtex 7 FPGA platform. Simulations, verification, and synthesis are performed on Vivado tools. Overall the proposed system provides an efficient and reliable solution for direction finding, pulse parameter estimation and tracking multiple signals simultaneously which is crucial in various applications. The achieved accuracy in the direction of arrival estimation and the promising results obtained from the FPGA implementation showcase the potential to use it in radar-based applications such as defense, surveillance, and Electronic intelligence (ELINT) systems.

Abstract

This thesis, titled “Exploring Financial Market Dynamics: A Simulation of Specu- lation Games with Information Cascades and Herding Effects Using Self-Organized Criticality,” delves into the complexities of decision-making processes in financial markets. The study underscores how information cascades and herding behaviors significantly influence individual and collective decisions, leading to market phenom- ena that are often unpredictable and volatile. In the literature review, an exhaustive examination of agent-based models (ABMs) in financial markets is conducted, with a particular emphasis on those based on the Minority Game. These models serve as foundational tools in understanding how decentralized decision-making can lead to emergent market patterns. Among the discussed models is the Speculation Game developed by Katahira, which, despite its robustness in simulating market dynamics, does not adequately account for herding behaviors—a critical aspect of real-world financial markets. To address this gap, this thesis proposes the SGIC model (Speculation Game with Information Cascades), an extension of Katahira’s model. In the SGIC model, by making the use of the concept of Self-Organized Criticality, information cascade and herding have been implemented. Self Organized Criticality is characterized the feature that small perturbation can result into avalanches when the system is in critical state. Such type of phenomenon are also observed in financial systems. The time series generated by the SGIC model has been validated by using Aug- mented Dickey-Fuller (ADF) unit root test for stationarity as well as for the pres- ence of stylized facts. It has been observed that SGIC model was able to successfully reproduce the stylized facts generally present in financial time-seriese. Furthermore, an in-depth analysis of herding patterns within the simulated time series reveals that these patterns bear a striking resemblance to those observed in real-world financial data, specifically from the BSE Sensex. This correlation underscores the model’s efficacy in capturing the nuances of herding behavior and its impact on market dynamics.

Abstract

Absolute separable (AS) quantum states are those states from which it is impossible to create entanglement, even under global unitary operations. It is known from the resource theory of non-absolute separability that the set of absolute separable states forms a convex and compact set, and global unitaries are free operations. We show that the action of a quantum switch controlled by an ancilla qubit over the global unitaries can break this robustness of AS states and produce ordinary separable states. First, we consider bipartite qubit systems and find the effect of quantum switch starting from the states sitting on the boundary of the set of absolute separable states. As particular examples, we illustrate what happens to modified Werner states and Bell diagonal (BD) states. For the Bell diagonal states, we provide the structure for the set of AS BD states and show how the structure changes under the influence of a switch. Further, we consider numerical generalization of the global unitary operations and show that it is always possible to take AS states out of the convex set under switching operations. We also generalized our results in higher dimensions. Several recent studies have demonstrated the utility of the quantum SWITCH as an important resource for enhancing the performance of various information processing tasks. In a quantum SWITCH, the advantages appear significantly due to the coherent superposition of alternative configurations of the quantum components which are controlled by an additional control system. Here we explore the impact of increasing the Hilbert-space dimension of the control system on the performance of the quantum SWITCH. In particular, we focus on a quantifier of the quantum SWITCH through the emergence of non-Markovianitxy and explicitly study their behavior when we increase the Hilbert-space dimension of the control system. We observe that increasing the Hilbert-space dimension of the control system leads to the corresponding enhancement of the non-Markovian memory induced by it. Our study demonstrates how the dimension of the control system can be harnessed to improve the quantum SWITCH-based information processing or communication tasks

Abstract

Remote labs are a technological breakthrough for accessing scientific knowledge, pushing the boundaries of traditional approaches to practical learning. They seek to provide students with unrestricted access to laboratory education and experimentation. Remote labs further prove invaluable in today’s times when pandemics such as COVID-19, geopolitical crises, and climatic conditions constantly threaten students’ access to practical learning in their institutions. Remote labs in such circumstances allow students remote access to practical equipment and laboratory resources. However, most of the developed remote labs today, either run on private networks, are specific to certain science domains, are too costly to replicate or require software installations, all of which hinder universal access to knowledge and education. With this motivation, the work presented in this thesis focuses on building a scalable software framework for real-time remote experimentation for Internet of Things (IoT)-based remote labs with lowlatency video streaming. This thesis is structured to comprise two parts. First, it explores the components and sequence flow of the developed end-to-end software architecture, emphasising the key design decisions for implementing the software solution and video streaming. It also discusses the introduction of an interoperability layer to support multiple IoT platforms and the implementation of a peer-to-peer (P2P) video streaming. The work further evaluates the developed architecture for its latency and capability of handling larger loads as the system scales. Secondly, the thesis also proposes the approach of multi-user multiplexing to address the scalability of remote labs hosting science experiments with near-instantaneous outputs based on electric currents, such as Kirchhoff’s laws. Such experiments and hardware nodes present a unique case compared to other practical experiments involving mechanical actuation. Multi-user multiplexing aims to provide users with concurrent access to the hardware nodes without impacting other users’ experience. The architecture proposed earlier can be modified slightly and adapted using WebSockets to facilitate the implementation of multi-user multiplexing. The modified WebSocket-based architecture constitutes three layers in contrast to the proposed four-layer architecture, and the two have been compared for their performance on latency and round trip time as part of this work

Abstract

A document has information emerging out of the existence of various physical entities or regions such as headings, paragraphs, figures, captions, tables, and backgrounds along with the textual content. To decipher the information in a document, a human reader uses a variety of additional cues, such as context, conventions, and information about language, script, location, and a complex reasoning process. Millions of documents are created and distributed daily over the Internet and printed media. Understanding, analyzing, sorting, and comparing a massive collection of documents in a limited time is a hectic job for humans. Here, the automatic document image understanding systems (DIUS) help humans do this tedious task within a limited time. The DIUS have typically a document image layout segmentation module and information extraction modules. This thesis focuses on the challenging problems related to document image layout segmentation in various types of documents and their applications. In this thesis, first, we analyse the various document images using deep features. The deep features are the features extracted using a pretrained deep neural network. To study deep texture features, we propose a deep network architecture that independently learns texture patterns, discriminative patches, and shapes to solve various document image analysis tasks. The considered tasks are document image classification, genre identification from book covers, scientific document figure classification, and script identification. The presented network learns global, texture, and discriminative features and combines them judicially based on the nature of the problems. We compare the performance of the proposed approach with state-of-the-art techniques on multiple publicly available datasets such as Book covers, RVL-CDIP, CVSI and DocFigure. Experiments show that our approach obtains significantly better performance over state-of-the-art for all tasks. Next, we focus on the problem of document image layout segmentation and propose a solution for a class of document images, including historical, scientific, and classroom slide document images. The historical document image segmentation problem is modeled as a pixel labeling task where each pixel in the document image is classified into one of the predefined labels, such as text, comment, decoration, and background. The method first extracts deep features from the superpixels of the document image. Then, we learn SVM classifier using these features and segment the document image. The pyramid pooling module is used to extract the logical regions of scientific document images. In the classroom slide images, the logical regions are distributed based on the location of the image. To utilize the location of the logical regions for slide image segmentation, we propose the architecture, Classroom Slide Segmentation Network (CSSN). The unique attributes of this architecture differ from most other semantic segmentation networks. We validate the performance of our segmentation architecture using publicly available benchmark datasets. Next, we analyze the output regions of document layout segmentation. Figures used in the documents are the complex regions to decipher the information by a DIUS. Hence, document figure classification (DFC) is an important stage of the DIUS. The design of a DFC system required well-defined figure categories and datasets. Existing datasets related to the classification of figures in the document images are limited in size and category. We introduce a scientific figure classification dataset named DocFigure. The dataset consists of 33K annotated figures of 28 different categories present in the document images, which correspond to scientific articles published in last several years. Manual annotation of such a large number (33K) of figures is time-consuming and cost-ineffective. We design a web-based annotation tool that can efficiently assign category labels to many figures with the minimum effort of human annotators. To benchmark our generated dataset on the classification task, we propose three baseline classification techniques using the deep feature, deep texture feature, and both. Our analysis found that the combination of both deep and texture features is more effective for document figure classification tasks than individual features. Finally, we propose the application backed by the research of this thesis. Slide presentations are an effective and efficient tool for classroom communication used by the teaching community. However, this teaching model can be challenging for blind and visually impaired (VI) students as such students require personal human assistance to understand the presented slide. This shortcoming motivates us to design a Classroom Slide Narration System (CSNS) that generates audio descriptions corresponding to the slide content. This problem poses an image-to-markup language generation task. Ext

Abstract

Unlike the classical scenario, conditional entropy in quantum systems can be negative. Such systems play a pivotal role in information-processing tasks such as super dense coding, one-way entanglement distillation, and state merging, to name a few. Quantum conditional entropy’s central role led to the formulation of a resource theoretic study, establishing the notion of free and resource states and a char- acterization of free operations. In this thesis, we delve into the study of quantum channels that affect the negative conditional entropy of a quantum system. The main focus of this thesis is the characterization of these channels from topological and information- theoretic perspectives. We consider two classes of quantum channels - Negative Conditional Entropy Breaking (NCEB) channels and Negative Conditional Entropy Annihilating (NCEA) channels. We de- fine such channels and examine their properties when combined serially. We also consider the property of parallel combinations of NCEB channels. Our exploration extends to complementary channels of NCEB, whereby we discover the information-leaking nature of NCEB channels. We exemplify NCEB and NCEA channels through standard depolarizing channels and further the characterization. Specif- ically for NCEA channels, we establish a sufficiency condition for global depolarizing channels and transpose depolarizing channels. Following this, we establish the relationship of NCEB and NCEA with newly introduced classes of quantum channels, such as Coherent Information Breaking (CIB) and Mutual Information Breaking (MIB), along with standard channels like zero capacity and entanglement breaking channels. Finally, we consider the perspective of maintaining negative conditional entropy in quantum states, which involves the identification of non-NCEB or non-NCEA channels. In this regard, we provide prescriptions to detect channels that do not affect the negativity of conditional entropy.