Abstract

Power, Performance, and Area have been some of the most crucial specification metrics to adjudge the quality of systems. Circuits based on nanoscale transistors operate at lower values of threshold and supply voltages, thereby providing better power-delay specifications. CMOS based nanoscale circuits (≤45nm) show higher static leakage losses than dynamic losses and become highly sensitive to PVT (Process parameters, Supply Voltage and Temperature) variations. Reliability analysis of the circuit in terms of propagation delays are greatly affected factors such as Bias Temperature Instability (BTI) and Hot Carrier Injection. Hence optimization is one of the most important aspects of the design flow. This thesis suggests an algorithm-based transistor sizing approach toward low-power and high-performance optimization of digital circuits. It proposes the use of 2 swarm intelligence based algorithms- Spider Monkey Optimization and Elephant Herding Optimization for transistor resizing in gates/ blocks either for low power or for high performance scenarios. In the first phase discussions, basic digital cells such as logic gates and blocks like Full Adders are optimized by resizing their transistors to meet the required specifications. The second phase involves the optimization of complex circuits comprising several of such basic cells/ blocks. This includes factors such as the critical signal path performance specifications, pairwise driver(s) and load(s) analysis, etc. This thesis proposes a python based top level module that scans through the circuit for potential critical/ near critical signals paths, identifies the types of basic cells present and the loads they drive, optimizes and finally places the appropriate resized cells to improve the overall circuit performances. We could obtain about 65% power and 44% in high performance optimization. Approximate computation deals in fields where the accuracy has a more relaxed tolerance level such asimage processing,speech processing etc. Thisis due to the fact that human senses are cannot perceive minute errors. We have utilized this performance-accuracy trade-off phenomenon to design low power systems with the help of algorithms. We have generated low power Full Adder designs through transistor pruning (selective removal of transistors) for speech processing systems with 67% power optimization. Accuracy has been tested and our design show improved performances as compared to previous works.

Abstract

Urban air quality monitoring faces a fundamental scalability crisis. With only 800 monitoring stations serving over 4,000 Indian cities, traditional sensor networks cannot capture pollution variations that occur every 300-500 meters in complex urban environments. The sparse coverage, combined with deployment costs of INR 40 lakhs to 1.6 crores per station, renders comprehensive monitoring economically unfeasible. Image-based air quality estimation emerges as a transformative alternative, yet existing approaches are constrained by limited datasets and poor generalization across cities. This thesis introduces two contributions that establish a new paradigm for scalable air quality monitoring. TRAQID (Traffic-Related Air Quality Image Dataset) provides the first comprehensive multiview dataset with 26,678 synchronized front-rear traffic image pairs, co-located environmental sensors, and systematic coverage across three seasons in Indian urban environments. Benchmark evaluation reveals existing state-of-the-art methods achieve only 75% accuracy on TRAQID, exposing critical limitations in current approaches. AQIFormer, our transformer-based architecture, presents image-based air quality estimation through weather-aware attention mechanisms, adaptive dual-view fusion, and multi-task temporal learning. The architecture achieves 89.96% accuracy on TRAQID—a remarkable 14.96% improvement over existing methods. Most significantly, AQIFormer demonstrates unprecedented cross-city generalization, maintaining 81.67% accuracy on independent Nagpur data with minimal adaptation (5% local samples), exhibiting only 8.29% performance degradation compared to typical 30%+ drops in existing approaches. Comprehensive analysis validates each innovation: dual-view integration contributes 11.46% improvement, multi-task learning enhances discriminability by 11.57%, and attention visualization reveals consistent focus on physically meaningful pollution sources across diverse urban environments. The architecture processes standard traffic imagery in real-time, enabling integration with existing camera infrastructure without specialized hardware. This research establishes the practical foundation for city-scale air quality monitoring at unprecedented spatial resolution and dramatically reduced costs. The robust cross-city performance enables immediate deployment across multiple urban environments, transforming air quality monitoring from expensive, sparse sensor networks to comprehensive, camera-based systems that can protect public health through enhanced environmental awareness.

Abstract

Understanding how narrative flow reflects cognitive processes such as memory, schema use, and event organization has long been a goal of cognitive science. Recent advancements in natural language processing, particularly the rise of large language models (LLMs), have introduced new tools for modeling and quantifying linguistic phenomena. Among these, a measure known as sequentiality—proposed by Sap et al. (2022)—aims to capture the narrative coherence of a story using LLM-derived probabilities. However, like many LLM-based approaches, its validity as a cognitively meaningful measure remains underexamined, especially in relation to human-annotated data. This thesis begins with a conceptual replication of the original sequentiality study, using a newly constructed dataset of paired autobiographical and biographical narratives written by the same individuals. Despite theoretical expectations of greater narrative structure in biographies, we fail to replicate the significant differences in sequentiality originally reported. To investigate further, we replicate the original analysis on Sap et al.’s dataset and identify a methodological bias stemming from topic-based components in the formulation. Removing this source of bias reduces the observed difference between narrative types, calling into question the suitability of the original metric for capturing narrative coherence. In the second part of this thesis, we propose a rectified version of the sequentiality measure, excluding the topic term and relying solely on sentence-to-sentence contextual prediction. To empirically validate this variant, we turn to the domain of Automated Essay Scoring (AES), where essays are annotated by human raters across multiple dimensions, including coherence. Using two well-established AES datasets, we demonstrate that the context-only sequentiality score aligns more closely with human judgments of narrative flow than the original formulation. Furthermore, we show that this measure improves the performance of AES systems when incorporated as a feature, highlighting its practical utility as an interpretable and cognitively motivated coherence metric. Together, these findings contribute to the growing literature on using LLMs as models of human linguistic behavior, while emphasizing the necessity for rigorous validation and methodological transparency when deploying such models in cognitive research. This work offers a clearer path toward integrating LLM-derived measures into applied NLP systems and opens avenues for future exploration in interpretability, bias mitigation, and discourse-level evaluation.

Abstract

Gynecologic and breast cancers are among the leading causes of morbidity and mortality in women worldwide, posing a significant threat to womens health. It is estimated that roughly one in five women will be diagnosed with cancer during their lifetime, with approximately one in twelve succumbing to the disease. Their molecular and cellular heterogeneity significantly contributes to therapeutic resistance and variability in clinical outcomes. Deciphering these complexities at various resolution is essential for advancing precision oncology. This thesis aims to unravel the molecular landscape of these cancers by leveraging bulk omics, single-cell RNA-seq (scRNA-seq), and spatial transcriptomics datasets. Metabolic reprogramming is a hallmark of cancer and is characterized by significant heterogeneity, presenting challenges for treatment efficacy and patient outcomes. Understanding metabolic adaptation and its heterogeneity in tumor tissues may provide new insights and help in developing personalized therapeutic strategies. We investigated the metabolic alterations of endometrial cancer (EC) to understand the variations in metabolism within tumor samples. Integration of transcriptomics data of EC (RNA-Seq) and the human genome-scale metabolic network was performed to identify the metabolic subtypes of EC and uncover the underlying dysregulated metabolic pathways and reporter metabolites in each subtype. The relationship between metabolic subtypes and clinical variables was explored. Further, we correlated the metabolic changes occurring at the transcriptome level with the genomic alterations. Based on metabolic profile, EC patients were stratified into two subtypes (metabolic subtype-1 and subtype-2) that significantly correlated to patient survival, tumor stages, mutation, and copy number variations. We observed the co-activation of the pentose phosphate pathway, one-carbon metabolism, and genes involved in controlling estrogen levels in metabolic subtype-2, which is linked to poor survival. PNMT and ERBB2 are also upregulated in metabolic subtype-2 samples and present on the same chromosome locus 17q12, which is amplified. PTEN and TP53 mutations show mutually exclusive behavior between subtypes and display a difference in survival. This work identifies metabolic subtypes with distinct characteristics at the transcriptome and genome levels, highlighting the metabolic heterogeneity within EC.

Abstract

Recent efforts to embed ethical reasoning into AI and machine learning systems have focused on training deep learning models to generate moral judgments in response to situational inputs. Parallel developments in the criminal justice system illustrate the application of such technologies, where risk assessment tools are employed to predict defendants’ risk scores. These scores are then used by legal institutions to make decisions related to parole, policing, and sentencing. Together, these approaches reflect a growing trend towards integrating AI into moral decision-making and as a potential replacement for human decision-making. In our work, we critically evaluate moral decision-making in AI systems through a philosophical lens, with particular emphasis on virtue ethics. First, we evaluate Delphi (Jiang et. al., 2021 [1]), which attempts to use the Rawlsian decision procedure as a bottom-up approach to capture abstract moral principles from crowdsourcing moral opinions across a diverse demographic. We critically evaluate this model on the basis of the effectiveness of (a) the current bottom-up approach of incorporating Rawlsian decision procedure, and (b) the proposed hybridization of the former with a top-down approach leading to reflective equilibrium. We argue against (a) on the grounds that the system fails to satisfy the criteria for being a competent moral judge as outlined by Rawls (Rawls, 1971 [2]); further, we argue against (b) on the grounds that reflective equilibrium requires phronesis or practical wisdom, which cannot be embedded into AI systems. We critically evaluate (b) in the light of Irikefe’s neo-Aristotelian, agentbased account of reflective equilibrium (Irikefe, 2019 [3]) to argue against the possibility of AI-ML systems to incorporate phronesis in order to attain reflective equilibrium. We then evaluate COMPAS, a risk assessment software that predicts the risk of recidivism of defendants based on responses to a questionnaire. We investigate COMPAS from a virtue ethics perspective by (1) discussing how punishment in the criminal justice system can be implemented as a practice, as defined within the framework of MacIntyre’s virtue ethics (MacIntyre, 2007 [4]) and (2) analyzing why such a practice cannot be implemented by tools like COMPAS.

Abstract

A control system aims to regulate the behavior of a dynamical system to achieve a desired output by adjusting the input. The control system design specifications for dynamical systems are typically provided in terms of the desired transient response and steady-state response. Performance measures like Rise Time, Settling Time, Steady-State Error, Control Effort, etc., help us understand the effectiveness of the designed controller in real-time. In traditional approaches, based on these specifications, we select design methods such as Root Locus or Bode Plot analysis, and algorithms like ProportionalIntegral-Derivative (PID), Linear Quadratic Regulator (LQR), and Model Predictive Control (MPC). After tuning with the allowed constraints, we may end up with a control system that meets our specifications. However, mathematical models of many real-world systems are not fully known. Meeting design specifications for such unknown dynamical systems is a challenging task. Partial or approximate system models often lead to inaccuracies due to internal modeling errors that are not always avoidable or correctable. Therefore, there is a need to explore control strategies that do not depend on mathematical models. Model-free Reinforcement Learning (RL) methods have demonstrated strong performance in this context. Their performance was comparable to traditional control design methods with system knowledge, and sometimes even better. This is due to their ability to learn optimal policies through interaction with the environment and adaptability to changes in real-time. Techniques such as Deep Q-Network (DQN), Proximal Policy Optimization (PPO), and Twin Delay Deep Deterministic Policy Gradient (TD3) have shown significant potential in handling complex control tasks without explicit system models. This thesis presents a new approach for designing controllers for dynamical systems with unknown mathematical models using a model-free RL framework based on the TD3 algorithm. The proposed method enables an RL agent to learn and shape the error dynamics and overall system response in trajectory tracking, particularly in state regulation tasks, through a suitably designed reward function. The effectiveness of the approach is validated on first-order and second-order linear systems, as well as a nonlinear pendulum system. The resulting controller significantly influences both the transient and steady-state behavior of an unknown system while adhering to control and input constraints. By systematically modifying the reward parameters, we analyze how reward shaping during training affects test-time performance metrics. This analysis establishes the relationship between reward design and control performance, providing a means to tune reward parameters to meet desired specifications. Furthermore, we introduce a data-driven optimization framework using Particle Swarm Optimization (PSO) to identify sets of reward parameters that satisfy given performance requirements. The method achieves exponential convergence of tracking error dynamics in linear systems without requiring explicit knowledge of system models.

Abstract

We present an efficient framework for identity-preserving face swapping into scenic portrait sketches. First, we analyze latent identity representations via a StyleGAN-based face cartoonization pipeline. Our analysis reveals domain-dependent shifts in StyleGAN’s identity encodings when mapping photos to sketches, underscoring the need for cross-domain consistency. Second, we introduce Portrait Sketching StyleGAN (PS-StyleGAN), a novel GAN architecture with attentive affine-transform blocks. These blocks learn to modulate a pretrained StyleGAN’s style codes by joint content-and-style attention, thereby learning style-specific identity transformations. PS-StyleGAN requires only a few photo-sketch pairs (and short training) to learn each style, making it broadly adaptable. Thus, PS-StyleGAN produces editable, expressive portrait sketches that faithfully preserve the input identity. Third, we extend diffusion-based generative modeling by adapting InstantID for sketch synthesis. Our diffusion module integrates strong semantic identity embeddings and landmark guidance (inspired by InstantID’s IdentityNet) to steer high-fidelity portrait generation. This yields personalized sketches with markedly improved identity fidelity while requiring only zero-shot (tuning-free) personalization. Finally, motivated by recent diffusion-based face-swapping advances, we build a real-time end-to-end pipeline. By encoding facial identity and pose as conditioning inputs and optimizing the diffusion process for speed, our system can seamlessly embed user faces into tourist-style sketches on the fly. Extensive evaluations on standard benchmarks confirm that our methods significantly enhance identity preservation and visual quality compared to prior approaches, advancing the state of the art in generative portrait stylization and face swapping.

Abstract

River water quality is fundamentally influenced by hydrological factors such as discharge and water temperature, both of which are increasingly affected by climate variability and human-induced pressures. These variables, when considered independently, can provide a limited view of water quality risks. However, compound events, where low discharge coincides with high water temperatures, can pose significant ecological threats, particularly to sensitive aquatic species and overall river health. Existing studies often rely on univariate or deterministic models, which fail to capture the joint behaviour of these interconnected variables. There is also a noticeable gap in translating advanced statistical models into practical tools for river water quality risk management, especially in computing compound event probabilities and supporting basin-specific decision-making. This study addresses these gaps by developing a copula-based joint modelling framework to quantify compound low-flow and high-temperature risks across six diverse Indian rivers: Yamuna, Bhadra, Kaveri, Mahi, Sabarmati, and Vardha. The study begins by fitting appropriate univariate probability distributions to river discharge and water temperatures using diagnostic tools such as skewness, kurtosis, Q-Q plots, and information criteria to ensure statistically sound marginal modelling. Dependence structures between the variables are assessed using rank correlation metrics to identify rivers where joint modelling is both appropriate and necessary. An extensive set of 18 copula families, including symmetric, asymmetric, and tail-dependent models, is evaluated to capture complex interdependencies across river basins. The selected copulas are then applied to estimate joint probabilities, conditional probabilities, and return periods of compound events of discharge and water temperatures of various river gauging stations of India. The findings reveal clear differences in compound event frequencies and intensities across the studied rivers, underscoring the need for localized and river-specific water quality management strategies. The key objectives of this study include identifying the best-fit univariate distributions, selecting the most appropriate copula families for joint modelling, and constructing a flexible, transferable statistical framework that can support future applications, including multivariable risk assessments, climate change impact studies, and adaptive water resource management. By addressing the underutilization of joint probabilistic frameworks for Indian river systems and demonstrating their practical application, this study contributes to a more holistic understanding of compound water quality risks and provides valuable guidance for sustainable river management and policy development

Abstract

Human knowledge is predominantly encoded in natural language. Automatic understanding of natural language has been a most awaited goal of AI/ML. Among several challenges of natural language processing and understanding, Machine Translation (MT) stands to be a major task. MT is the process of converting the source text in one natural language to another natural language. With the advent of neural models, the field of neural MT (NMT) has made significant progress, yet the challenge of effectively evaluating the translation quality remains critical. This thesis focuses on the automatic evaluation of machine translation outputs, addressing both the limitations of existing metrics and proposing innovative approaches to enhance evaluation accuracy. Existing widely used metrics, such as BLEU and METEOR, are scrutinized for their reliance on n-gram overlap and limited linguistic insight, often failing to capture nuances of language such as fluency, adequacy, and semantic equivalence. To mitigate these shortcomings, this research explores advanced evaluation methods leveraging pre-trained multilingual embeddings and deep learning, which model linguistic phenomena more comprehensively. This thesis introduces several novel automatic evaluation approaches (unsupervised referencebased, unsupervised reference-free and supervised approaches) that integrates semantic similarity and contextual relevance by employing pre-trained language models. Through extensive experimentation, our metrics are benchmarked against the traditional ones, demonstrating superior performance in correlating with human judgments. Additionally, this work investigates the existing evaluation datasets and evaluation approaches, discusses their limitations, and also provides decision trees to help the researchers choose particular evaluation criteria or a metric based on available computation and linguistic resources. Ultimately, this thesis contributes to the field by offering robust methodologies for MT output evaluation, emphasizing the need for metrics that align more closely with human perceptions of translation quality. The proposed advancements pave the way for more reliable and insightful evaluation systems, fostering continued innovation and improvement in machine translation technologies.

Abstract

Power transmission infrastructure requires regular inspection and maintenance to ensure reliable operation, yet traditional methods involving human crews pose significant safety risks and operational costs, while autonomous aerial inspection systems face limitations in achieving closecontact inspection due to collision avoidance requirements and existing wire-traversing robots struggle with obstacle navigation and complex deployment procedures. This thesis presents WireFlie, a novel hybrid robotic system that addresses these challenges by integrating aerial mobility with contact-based wire traversal capabilities through a dual-arm underactuated mechanism mounted on a multi-rotor drone platform, enabling seamless transitions between flight and rolling locomotion modes via a sequential two-phase operation for safe wire engagement. The system employs computer vision for autonomous wire detection and obstacle classification, coupled with hierarchical control for precise wire approach and an obstacle avoidance module that automatically selects appropriate operational modes based on detected obstacle types, enabling dual-arm rolling over common obstacles, single-arm rolling for spacer configurations, and flight mode transitions for larger obstacles like transmission pylons. Experimental validation demonstrates the system’s effectiveness across multiple scenarios, confirming stable operation on inclined transmission lines, consistent autonomous wire engagement, and successful obstacle navigation through various configurations, with real-world validation on representative transmission line setups demonstrating the complete operational sequence from autonomous wire detection through obstacle navigation to controlled detachment. This research contributes to robotic infrastructure inspection by demonstrating the feasibility of hybrid aerial-ground locomotion for power line monitoring, establishing a foundation for next-generation transmission line inspection systems that combine the mobility advantages of aerial platforms with the precision requirements of contact-based inspection, ultimately enhancing power grid maintenance capabilities while improving operational safety and efficiency. Keywords: Transmission line inspection, underactuated mechanisms, hybrid locomotion, obstacle avoidance, autonomous control, robotic infrastructure monitoring