ScholarSphere: IITJ Research Insights Hub

Building high-fidelity 3D maps from LiDAR scans in dynamic environments remains a challenging task due to ghosting artifacts caused by transient objects and geometric inconsistencies arising from sparse spatial observations. In this work, we propose a 4D implicit neural mapping framework that integrates localized temporal modeling with a structurally decoupled static–dynamic representation. We employed compactly supported Wendland radial basis functions as temporal bases to enforce strict temporal locality and eliminate long-range temporal leakage inherent in global basis functions such as the discrete cosine basis. Then, we propose a spatial context module that refines multi-resolution hash-encoded features through a residual multilayer perceptron to improve geometric coherence in sparsely observed regions. Finally, we proposed a dual-path decoder that separates time-invariant static geometry from time-variant dynamic deformation via independent prediction heads sharing a common backbone, enabling clean static map extraction via simple geometric thresholding on the decoupled representation. The proposed framework is trained end to end in an unsupervised manner. We evaluate our approach on CoFusion, Newer College, and KTH Dynamic Map Benchmark datasets and demonstrate improved reconstruction fidelity and dynamic segmentation accuracy over existing methods. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2026.

Satellite-based free-space quantum key distribution (QKD) provides a practical framework for achieving secure global communication beyond the limitations of optical fibers. In this work, the quantum bit error rate (QBER) and secure key rate of four representative protocols-BB84, B92, BBM92, and E91 are investigated over low earth orbit (LEO) links in both uplink and downlink configurations. The optical link is modeled using a Gaussian beam formalism, incorporating the effects of diffraction, pointing errors, atmospheric turbulence, and background noise contributions. The protocols are examined under day and night-time operating conditions, and their dependence on the zenith angle is analyzed. The findings show that downlink links generally exhibit lower QBER and higher secure key rates than uplinks, and among prepare-and-measure schemes, BB84 consistently outperforms B92, while in entanglement-based approaches, BBM92 achieves higher key rates than E91. © 2026 World Scientific Publishing Company.

Citywide incidents such as crimes, accidents, and public safety threats contribute to substantial societal disruption and economic loss. Accurate prediction of such incidents can significantly aid city administrators in proactive response planning. Existing approaches model the incident prediction as a spatio-temporaltask, but often neglect the inter-region spatial long-tailed distribution of incidents. This uneven distribution introduces spatial bias in learning which causes models to overfit regions with frequent incidents (head regions) while underfit the regions with occasional incidents (tail regions). Furthermore, model learning is hindered by intra-region category imbalance, where certain incident types (e.g., theft) dominate over rarer categories (e.g., robbery) within the same region. To address inter and intra region challenges, we propose an approach named SLIP (Spatial Long-tail Incident Prediction). Specifically, for inter-region skewness, SLIP adopts a multi-expert design comprising a common feature extraction backbone followed by three expert branches. In addition, to mitigate the intra-region category imbalance, we utilizes a variant of focal loss, particularly for positive-negative imbalance. SLIP outperforms spatio-temporal state-of-the-art methods by 2-11% in Macro F1, 4-11% in Micro F1, and 1-11% in Severity Weighted F1 across Los Angeles and Chicago cities for the urban crime dataset. Additionally, we incorporate fairness metrics into the evaluation and present a comprehensive comparison of spatio-temporal incident prediction. © 2015 IEEE.

Organic molecule-based memristive devices are promising candidates for next-generation data storage devices due to their scalability and low cost. This work discusses a resistive memory device based on a small organic molecule 2-mercapto-1-methylimidazole (MMI) and the polymer poly 4-vinylpyridine (PVP), which exhibits stable resistive switching with a high on–off ratio (5.48 × 103), retention over 3.6 × 104 s and endurance over 700 cycles. In addition to memory behavior, the MMI organic molecule-based memristor exhibits synaptic functions crucial for neuromorphic computing. The device shows analog modulation of conductance in accordance with voltage pulse protocols, mimicking essential biological learning mechanisms, including potentiation, depression, short-term plasticity (STP), long-term plasticity (LTP), and paired-pulse facilitation (PPF). The device also demonstrates associative learning via Pavlovian conditioning, demonstrating its potential as a hardware-implemented artificial synapse in emerging brain-inspired systems. The carrier transport in these devices follows multiple conduction mechanisms including trap-free and trap-assisted space-charge limited conduction (SCLC), and Ohmic conduction. The switching is attributed to metallic filament formation from the top electrode which is further supported by impedance measurements. This study highlights the potential of organic molecular memristors for next-generation memory and neuromorphic computing. This journal is © The Royal Society of Chemistry, 2026.

Neurodegenerative disorders such as Parkinson's (PD) and Alzheimer's diseases (AD) are linked with the assembly and accumulation of proteins into structured scaffold called amyloids. These diseases pose significant challenges due to their complex and multifaceted nature. While the primary focus has been on endogenous amyloids, recent evidence suggests that bacterial amyloids may contribute to the development and exacerbation of such disorders. The gut-brain axis is emerging as a communication pathway between bacterial and human amyloids. This review delves into the novel role and potential mechanism of bacterial amyloids in modulating human amyloid formation and the progression of AD and PD.

In the MaxSAT with Cardinality Constraint problem (CC-MaxSAT), we are given a CNF-formula Φ, and a positive integer k, and the goal is to find an assignment β with at most k variables set to true (also called a weight k-assignment) such that the number of clauses satisfied by β is maximized. Maximum Coverage can be seen as a special case of CC-MaxSat, where the formula Φ is monotone, i.e., does not contain any negative literals. CC-MaxSat and Maximum Coverage are extremely well-studied problems in the approximation algorithms as well as the parameterized complexity literature. Our first conceptual contribution is that CC-MaxSat and Maximum Coverage are equivalent to each other in the context of FPT-Approximation parameterized by k (here, the approximation is in terms of the number of clauses satisfied/elements covered). In particular, we give a randomized reduction from CC-MaxSat to Maximum Coverage running in time O(1/ϵ)k · (m + n)O(1) that preserves the approximation guarantee up to a factor of (1 − ϵ). Furthermore, this reduction also works in the presence of “fairness” constraints on the satisfied clauses, as well as matroid constraints on the set of variables that are assigned true. Here, the “fairness” constraints are modeled by partitioning the clauses of the formula Φ into r different colors, and the goal is to find an assignment that satisfies at least tj clauses of each color 1 ≤ j ≤ r. Armed with this reduction, we focus on designing FPT-Approximation schemes (FPT-ASes) for Maximum Coverage and its generalizations. Our algorithms are based on a novel combination of a variety of ideas, including a carefully designed probability distribution that exploits sparse coverage functions. These algorithms substantially generalize the results in Jain et al. [SODA 2023] for CC-MaxSat and Maximum Coverage for Kd,d-free set systems (i.e., no d sets share d elements), as well as a recent FPT-AS for Matroid Constrained Maximum Coverage by Sellier [ESA 2023] for frequency-d set systems.

The technique of micro-exfoliation has gained prominence as a highly effective and adaptable method for exploiting two-dimensional (2D) materials, such as graphene Transition metal dichalcogenides (TMDCs), Borophene, Molybdenum disulfide (MoS2), among others. This paper presents an analysis of optical images and the micro exfoliation technique, focusing on the application to MoS2 and graphene. Additionally, the study investigates the exfoliated sheet of graphene, MoS2, and their hybrid on a (111) crystal plane of silicon wafer. The micro-exfoliation technique employed for MoS2 involves a mechanical process that gently disentangles the layers of MoS2 from the larger crystal structure, resulting in the formation of ultrathin two-dimensional nanosheets. This paper comprehensively analyses the exfoliation processes' mechanisms, emphasizing the intricate relationship between van der Waals forces, interlayer bonding, and external forces. The micro-mechanical exfoliation, a fundamental technique, entails the utilization of adhesive scotch tape to remove monolayers from a large MoS2 crystal delicately. The integration of MoS2 into various applications such as electronics, optoelectronics, sensors, and energy storage devices has been driven by its exceptional properties, including its distinctive electronic, optical, and mechanical characteristics. Furthermore, the ability to adjust the bandgap of MoS2 has created novel opportunities for potential applications in the field of semiconductors. This paper provides a succinct summary of recent studies, that have concentrated on the optical characterization of MoS2 monolayers. Optical and Raman spectroscopy was employed to characterize the 2D sheets of MoS2 and its hybrid materials.