Searcharxiv⌕ Search

arXiv subjects

Aryan Singh

Publications and source records attributed to Aryan Singh.

16 recordsLinked to original sources
arXiv2026

Freezing of Gait Prediction Under Spatial Occlusion: An IMU-Supervised Cross-Modal Distillation Approach

Parkinson's disease is a progressive neurodegenerative disorder characterised by gradual deterioration of movement control. Automated freezing-of-gait (FOG) detection supports the objective assessment of gait-related motor impairment. Two common approaches are used for FOG prediction: (i) analysing video recordings of the patient's movements and (ii) analysing data collected using inertial measurement unit (IMU) wearable sensors attached to the patient's lower limbs. Video-based approaches may suffer detection errors during continuous turning-in-place tasks because the lower limbs undergo substantial geometric self-occlusion, degrading pose-estimation accuracy. IMU-based approaches are generally less affected by visual occlusion; however, they are difficult to deploy outside clinical or laboratory settings, as the sensors must be attached securely and remain in place throughout the assessment. Motivated by this, we propose a cross-modal subspace distillation framework to mitigate the limitations of unimodal FOG detection by combining IMU accuracy with video-based practicality. We extract invariant latent topologies from a pre-trained kinematic oracle to structurally supervise a non-encoded visual architecture during training. To resolve periods of severe spatial occlusion, a dual-stream visual model probabilistically fuses skeletal graph nodes and continuous spatial pixels, dynamically shifting reliance to uninterrupted pixel boundaries as joint tracking confidence drops. Evaluated against a public, multi-modal sequence dataset of Parkinson's individuals executing continuous $360^\circ$ turns, empirical results demonstrate that applying sensory boundary topologies strictly mitigates tracking evaluation entropy. Our constrained optimisation confirms that highly precise FOG prediction bounds can be achieved over zero-wearable inference environments.

cs.CV↗
arXiv2026

Supervised Cross-Modal Feature Alignment for Zero-Wearable Freezing of Gait Detection in Parkinsonism

Objective assessment of Freezing of Gait (FoG) in Parkinson's disease (PD) relies predominantly on wearable Inertial Measurement Units (IMUs). While IMUs provide optimal kinematic precision, mandatory sensor attachment restricts continuous clinical deployment. Conversely, unobtrusive vision-based alternatives suffer substantial classification errors during turning-in-place tasks, where geometric self-occlusion degrades deterministic skeletal coordinates and obscures the high-frequency precursors required for FoG detection. To resolve these physical observation limits, we propose a supervised cross-modal subspace distillation framework. During optimisation, pre-trained kinematic data from IMU sensors and contextual clinical metadata act as oracles to guide a deployable visual architecture. By incorporating joint velocity and acceleration derivatives, utilising a confidence-based gating mechanism, the visual model mitigates some of the tracking errors during occlusion events. Empirical evaluations confirm this latent alignment transfers the predictive fidelity of hardware sensors directly into the visual representation, yielding $85.5\%$ accuracy, and $82.4\%$ balanced accuracy. All the while maintaining a vision only model at inference.

cs.CV↗
arXiv2026

BOOSTEDSOSA: Accelerated Inferencing for Low Variance Stochastic Online Scheduling

Heterogeneous scheduling in stochastic, online envi- ronments, such as high-performance computing (HPC) systems, presents a significant challenge. Stochastic Online Scheduling Accelerators (SOSAs) offer a promising solution, but their effectiveness is compromised by a reliance on runtime estimates provided by users. These estimates introduce substantial vari- ance into the scheduling process (mean MAE in hundreds of Core-Days), thereby weakening the competitiveness of Stochastic Online Scheduling algorithms as their competitive-ratio bound increases with runtime variability. To address this limitation, we introduce BOOSTEDSOSA, a dual-FPGA ML-assisted Scheduling architecture that integrates a Machine Learning predictor for expected processing times, with a novel temporal-aware training policy. The predictor estimates job runtimes using only scheduler parameters available at submission time, enabling its use in existing HPC systems. Using historical real-world HPC job data (from the Argonne Leadership Comput- ing Facility, MIT Supercloud and UIUC Blue Waters workload datasets), we show that the predictor reduces MAE by up to 63.85% compared to user runtime estimates, and the additive training policy reduces MAE by up to 71.88% compared to a static model. End-to-end, BOOSTEDSOSA achieves an average 17x speedup over an AVX-optimized software baseline and processes up to 1,711 jobs/seconds

cs.AR↗
arXiv2026

Decomposition-based Energy-based Dual-Phase Dynamics Identification for Nonlinear MDOF Systems

System identification is an important step in modeling and evaluating vibrating structures, but many nonlinear system identification methods rely heavily on data-driven approaches that may not preserve physical consistency. This research extends the Energy-based Dual-phase Dynamics Identification (EDDI) method to multiple-degree-of-freedom (MDOF) mechanical structures undergoing nonlinear vibrations. The original EDDI framework was designed for single-degree-of-freedom (SDOF) systems and operates in two phases: the first identifies a model for internal nonconservative force, and the second captures internal conservative force. However, EDDI assumes that the potential energy is zero whenever the displacement is zero. For MDOF systems, this assumption requires all degrees of freedom (DOFs) to achieve zero displacement simultaneously, which simply occurs too infrequently in multimodal responses for direct application. To overcome this limitation, this work introduces Decomposition-based EDDI, which applies EDDI to decomposed response components to enable nonlinear system identification of MDOF systems. Wavelet-Bounded Empirical Mode Decomposition is used to extract nearly orthogonal, monochromatic intrinsic mode functions (IMFs) from the measured displacements. Each IMF is then treated as an individual SDOF oscillator and processed using EDDI to estimate its nonconservative and conservative internal forces. The IMF forces are then summed to reconstruct the total nonconservative and conservative forces acting on each physical DOF, which are used to identify the damping and stiffness models, respectively. The proposed EDDI framework is experimentally validated on a two-story tower structure with strong stiffness nonlinearity coupling the two floors. The results demonstrate the efficacy of EDDI in isolating and identifying complex, multi-modal nonlinear structural dynamics.

math.DS↗
arXiv2026

Determination of the best dark matter profile for the Milky Way with Gaia DR3 using Bayesian Model Comparison

We carry out a Bayesian model comparison analysis to determine the dark matter model that best describes the Milky Way rotation curve, using four recent compilations of rotation-curve velocities with the Gaia data. We compare these data using three different baryonic models along with seven different dark matter models. Within the adopted modeling framework and Gaia-based rotation curve datasets, we find that the Einasto profile provides the preferred phenomenological fit compared to the NFW profile} across most combinations of datasets and baryonic models. We also find that the cored dark profiles are preferred over the (cuspy) NFW profile. We also test MOND using three different interpolating functions and find that, within the implementations considered, the models provide poorer fits than the preferred dark matter profiles. Finally, among the different baryonic models considered, none is decisively favored over the others.

astro-ph.GA↗
arXiv2025

Cross-Model Verification of Wall-Bounded Flows using Finite-JAX

Accurate prediction of wall-bounded flows remains central to advancing both theoretical understanding and computational methods in fluid mechanics. In this study, we perform a numerical simulation of channel flow using a complementary approach: a high-performance, differentiable finite-difference solver developed in JAX (Finite-JAX), and an analytical solution derived from the Navier-Stokes Equations, also known as the Hagen-Poiseuille equation. The solver is applied to the incompressible Navier-Stokes equations, along with appropriate boundary conditions, to capture canonical flow features, including velocity profiles and pressure gradients. Cross-model verification is conducted by systematically comparing numerical results between Finite-JAX and the analytical solution, with a focus on velocity distributions. In addition, numerical results are benchmarked against analytical solutions for the laminar regime, allowing direct quantification of the verification accuracy. Our findings demonstrate that cross-model verification not only strengthens confidence in simulation fidelity but also provides a pathway for integrating differentiable solvers with established computational fluid dynamics platforms, paving the way for future fluid flow research. The performance of Finite-JAX on Wall-Bounded Flows is 0.014765 in the L2 norm.

physics.flu-dyn↗
arXiv2025

AI Based Font Pair Suggestion Modelling For Graphic Design

One of the key challenges of AI generated designs in Microsoft Designer is selecting the most contextually relevant and novel fonts for the design suggestions. Previous efforts involved manually mapping design intent to fonts. Though this was high quality, this method does not scale for a large number of fonts (3000+) and numerous user intents for graphic design. In this work we create font visual embeddings, a font stroke width algorithm, a font category to font mapping dataset, an LLM-based category utilization description and a lightweight, low latency knowledge-distilled mini language model (Mini LM V2) to recommend multiple pairs of contextual heading and subheading fonts for beautiful and intuitive designs. We also utilize a weighted scoring mechanism, nearest neighbor approach and stratified sampling to rank the font pairs and bring novelty to the predictions.

cs.CV↗
arXiv2025

Image Segmentation: Inducing graph-based learning

This study explores the potential of graph neural networks (GNNs) to enhance semantic segmentation across diverse image modalities. We evaluate the effectiveness of a novel GNN-based U-Net architecture on three distinct datasets: PascalVOC, a standard benchmark for natural image segmentation, WoodScape, a challenging dataset of fisheye images commonly used in autonomous driving, introducing significant geometric distortions; and ISIC2016, a dataset of dermoscopic images for skin lesion segmentation. We compare our proposed UNet-GNN model against established convolutional neural networks (CNNs) based segmentation models, including U-Net and U-Net++, as well as the transformer-based SwinUNet. Unlike these methods, which primarily rely on local convolutional operations or global self-attention, GNNs explicitly model relationships between image regions by constructing and operating on a graph representation of the image features. This approach allows the model to capture long-range dependencies and complex spatial relationships, which we hypothesize will be particularly beneficial for handling geometric distortions present in fisheye imagery and capturing intricate boundaries in medical images. Our analysis demonstrates the versatility of GNNs in addressing diverse segmentation challenges and highlights their potential to improve segmentation accuracy in various applications, including autonomous driving and medical image analysis.

cs.CV↗
arXiv2024

A Data-Driven, Energy-based Approach for Identifying Equations of Motion in Vibrating Structures Directly from Measurements

Determining the underlying equations of motion and parameter values for vibrating structures is of great concern in science and engineering. This work introduces a new data-driven approach called the energy-based dual-phase dynamics identification (EDDI) method for identifying the nonlinear dynamics of single-degree-of-freedom oscillators. The EDDI method leverages the energies of the system to identify the governing dynamics through the forces acting on the oscillator. The approach consists of two phases: a model-dissipative and model-stiffness identification. In the first phase, the fact that kinetic and mechanical energies are equivalent when the displacement is zero is leveraged to compute the energy dissipated and a corresponding model for the nonlinear damping of the system. In the second phase, the energy dissipated is used to compute the mechanical energy (ME), which is then used to obtain a reformulated Lagrangian. The conservative forces acting on the oscillator are then computed by taking the derivative the Lagrangian, then a model for the nonlinear stiffness is identified by solving a system of linear equations. The resulting governing equations are identified by including both the nonlinear damping and stiffness terms. A key novelty of the EDDI method is that the only thing required to perform the identification is free-response measurements and the mass of the oscillator. No prior understanding of the dynamics of the system is necessary to identify the underlying dynamics, such that the EDDI method is a truly data-driven method. The method is demonstrated using simulated and measured responses of nonlinear single-degree-of-freedom systems with a variety of nonlinear mechanisms.

math.DS↗
arXiv2024

Deformable Convolution Based Road Scene Semantic Segmentation of Fisheye Images in Autonomous Driving

This study investigates the effectiveness of modern Deformable Convolutional Neural Networks (DCNNs) for semantic segmentation tasks, particularly in autonomous driving scenarios with fisheye images. These images, providing a wide field of view, pose unique challenges for extracting spatial and geometric information due to dynamic changes in object attributes. Our experiments focus on segmenting the WoodScape fisheye image dataset into ten distinct classes, assessing the Deformable Networks' ability to capture intricate spatial relationships and improve segmentation accuracy. Additionally, we explore different loss functions to address class imbalance issues and compare the performance of conventional CNN architectures with Deformable Convolution-based CNNs, including Vanilla U-Net and Residual U-Net architectures. The significant improvement in mIoU score resulting from integrating Deformable CNNs demonstrates their effectiveness in handling the geometric distortions present in fisheye imagery, exceeding the performance of traditional CNN architectures. This underscores the significant role of Deformable convolution in enhancing semantic segmentation performance for fisheye imagery.

cs.CV↗
arXiv2024

Subgraph Clustering and Atom Learning for Improved Image Classification

In this study, we present the Graph Sub-Graph Network (GSN), a novel hybrid image classification model merging the strengths of Convolutional Neural Networks (CNNs) for feature extraction and Graph Neural Networks (GNNs) for structural modeling. GSN employs k-means clustering to group graph nodes into clusters, facilitating the creation of subgraphs. These subgraphs are then utilized to learn representative `atoms` for dictionary learning, enabling the identification of sparse, class-distinguishable features. This integrated approach is particularly relevant in domains like medical imaging, where discerning subtle feature differences is crucial for accurate classification. To evaluate the performance of our proposed GSN, we conducted experiments on benchmark datasets, including PascalVOC and HAM10000. Our results demonstrate the efficacy of our model in optimizing dictionary configurations across varied classes, which contributes to its effectiveness in medical classification tasks. This performance enhancement is primarily attributed to the integration of CNNs, GNNs, and graph learning techniques, which collectively improve the handling of datasets with limited labeled examples. Specifically, our experiments show that the model achieves a higher accuracy on benchmark datasets such as Pascal VOC and HAM10000 compared to conventional CNN approaches.

cs.CV↗
arXiv2024

Large Language Models aren't all that you need

This paper describes the architecture and systems built towards solving the SemEval 2023 Task 2: MultiCoNER II (Multilingual Complex Named Entity Recognition) [1]. We evaluate two approaches (a) a traditional Conditional Random Fields model and (b) a Large Language Model (LLM) fine-tuned with a customized head and compare the two approaches. The novel ideas explored are: 1) Decaying auxiliary loss (with residual) - where we train the model on an auxiliary task of Coarse-Grained NER and include this task as a part of the loss function 2) Triplet token blending - where we explore ways of blending the embeddings of neighboring tokens in the final NER layer prior to prediction 3) Task-optimal heads - where we explore a variety of custom heads and learning rates for the final layer of the LLM. We also explore multiple LLMs including GPT-3 and experiment with a variety of dropout and other hyperparameter settings before arriving at our final model which achieves micro & macro f1 of 0.85/0.84 (on dev) and 0.67/0.61 on the test data . We show that while pre-trained LLMs, by themselves, bring about a large improvement in scores as compared to traditional models, we also demonstrate that tangible improvements to the Macro-F1 score can be made by augmenting the LLM with additional feature/loss/model engineering techniques described above.

cs.CL↗
arXiv2023

Optimizing Ego Vehicle Trajectory Prediction: The Graph Enhancement Approach

Predicting the trajectory of an ego vehicle is a critical component of autonomous driving systems. Current state-of-the-art methods typically rely on Deep Neural Networks (DNNs) and sequential models to process front-view images for future trajectory prediction. However, these approaches often struggle with perspective issues affecting object features in the scene. To address this, we advocate for the use of Bird's Eye View (BEV) perspectives, which offer unique advantages in capturing spatial relationships and object homogeneity. In our work, we leverage Graph Neural Networks (GNNs) and positional encoding to represent objects in a BEV, achieving competitive performance compared to traditional DNN-based methods. While the BEV-based approach loses some detailed information inherent to front-view images, we balance this by enriching the BEV data by representing it as a graph where relationships between the objects in a scene are captured effectively.

cs.CV↗
arXiv2023

Connecting the Dots: Graph Neural Network Powered Ensemble and Classification of Medical Images

Deep learning models have demonstrated remarkable results for various computer vision tasks, including the realm of medical imaging. However, their application in the medical domain is limited due to the requirement for large amounts of training data, which can be both challenging and expensive to obtain. To mitigate this, pre-trained models have been fine-tuned on domain-specific data, but such an approach can suffer from inductive biases. Furthermore, deep learning models struggle to learn the relationship between spatially distant features and their importance, as convolution operations treat all pixels equally. Pioneering a novel solution to this challenge, we employ the Image Foresting Transform to optimally segment images into superpixels. These superpixels are subsequently transformed into graph-structured data, enabling the proficient extraction of features and modeling of relationships using Graph Neural Networks (GNNs). Our method harnesses an ensemble of three distinct GNN architectures to boost its robustness. In our evaluations targeting pneumonia classification, our methodology surpassed prevailing Deep Neural Networks (DNNs) in performance, all while drastically cutting down on the parameter count. This not only trims down the expenses tied to data but also accelerates training and minimizes bias. Consequently, our proposition offers a sturdy, economically viable, and scalable strategy for medical image classification, significantly diminishing dependency on extensive training data sets.

cs.CV↗
arXiv2023

Compact & Capable: Harnessing Graph Neural Networks and Edge Convolution for Medical Image Classification

Graph-based neural network models are gaining traction in the field of representation learning due to their ability to uncover latent topological relationships between entities that are otherwise challenging to identify. These models have been employed across a diverse range of domains, encompassing drug discovery, protein interactions, semantic segmentation, and fluid dynamics research. In this study, we investigate the potential of Graph Neural Networks (GNNs) for medical image classification. We introduce a novel model that combines GNNs and edge convolution, leveraging the interconnectedness of RGB channel feature values to strongly represent connections between crucial graph nodes. Our proposed model not only performs on par with state-of-the-art Deep Neural Networks (DNNs) but does so with 1000 times fewer parameters, resulting in reduced training time and data requirements. We compare our Graph Convolutional Neural Network (GCNN) to pre-trained DNNs for classifying MedMNIST dataset classes, revealing promising prospects for GNNs in medical image analysis. Our results also encourage further exploration of advanced graph-based models such as Graph Attention Networks (GAT) and Graph Auto-Encoders in the medical imaging domain. The proposed model yields more reliable, interpretable, and accurate outcomes for tasks like semantic segmentation and image classification compared to simpler GCNNs

eess.IV↗