arXiv subjects

Aditya Sharma

Publications and source records attributed to Aditya Sharma.

At least 37 records · Page 2Linked to original sources

arXiv2025

Weak Links in LinkedIn: Enhancing Fake Profile Detection in the Age of LLMs

Large Language Models (LLMs) have made it easier to create realistic fake profiles on platforms like LinkedIn. This poses a significant risk for text-based fake profile detectors. In this study, we evaluate the robustness of existing detectors against LLM-generated profiles. While highly effective in detecting manually created fake profiles (False Accept Rate: 6-7%), the existing detectors fail to identify GPT-generated profiles (False Accept Rate: 42-52%). We propose GPT-assisted adversarial training as a countermeasure, restoring the False Accept Rate to between 1-7% without impacting the False Reject Rates (0.5-2%). Ablation studies revealed that detectors trained on combined numerical and textual embeddings exhibit the highest robustness, followed by those using numerical-only embeddings, and lastly those using textual-only embeddings. Complementary analysis on the ability of prompt-based GPT-4Turbo and human evaluators affirms the need for robust automated detectors such as the one proposed in this study.

cs.SI↗

arXiv2025

Automated Defect Identification and Categorization in NDE 4.0 with the Application of Artificial Intelligence

This investigation attempts to create an automated framework for fault detection and organization for usage in contemporary radiography, as per NDE 4.0. The review's goals are to address the lack of information that is sufficiently explained, learn how to make the most of virtual defect increase, and determine whether the framework is viable by using NDE measurements. As its basic information source, the technique consists of compiling and categorizing 223 CR photographs of airplane welds. Information expansion systems, such as virtual defect increase and standard increase, are used to work on the preparation dataset. A modified U-net model is prepared using the improved data to produce semantic fault division veils. To assess the effectiveness of the model, NDE boundaries such as Case, estimating exactness, and misleading call rate are used. Tiny a90/95 characteristics, which provide strong differentiating evidence of flaws, reveal that the suggested approach achieves exceptional awareness in defect detection. Considering a 90/95, size error, and fake call rate in the weld area, the consolidated expansion approach clearly wins. Due to the framework's fast derivation speed, large images can be broken down efficiently and quickly. Professional controllers evaluate the transmitted system in the field and believe that it has a guarantee as a support device in the testing cycle, irrespective of particular equipment cut-off points and programming resemblance.

cs.CV↗

arXiv2025

MedHELM: Holistic Evaluation of Large Language Models for Medical Tasks

While large language models (LLMs) achieve near-perfect scores on medical licensing exams, these evaluations inadequately reflect the complexity and diversity of real-world clinical practice. We introduce MedHELM, an extensible evaluation framework for assessing LLM performance for medical tasks with three key contributions. First, a clinician-validated taxonomy spanning 5 categories, 22 subcategories, and 121 tasks developed with 29 clinicians. Second, a comprehensive benchmark suite comprising 35 benchmarks (17 existing, 18 newly formulated) providing complete coverage of all categories and subcategories in the taxonomy. Third, a systematic comparison of LLMs with improved evaluation methods (using an LLM-jury) and a cost-performance analysis. Evaluation of 9 frontier LLMs, using the 35 benchmarks, revealed significant performance variation. Advanced reasoning models (DeepSeek R1: 66% win-rate; o3-mini: 64% win-rate) demonstrated superior performance, though Claude 3.5 Sonnet achieved comparable results at 40% lower estimated computational cost. On a normalized accuracy scale (0-1), most models performed strongly in Clinical Note Generation (0.73-0.85) and Patient Communication & Education (0.78-0.83), moderately in Medical Research Assistance (0.65-0.75), and generally lower in Clinical Decision Support (0.56-0.72) and Administration & Workflow (0.53-0.63). Our LLM-jury evaluation method achieved good agreement with clinician ratings (ICC = 0.47), surpassing both average clinician-clinician agreement (ICC = 0.43) and automated baselines including ROUGE-L (0.36) and BERTScore-F1 (0.44). Claude 3.5 Sonnet achieved comparable performance to top models at lower estimated cost. These findings highlight the importance of real-world, task-specific evaluation for medical use of LLMs and provides an open source framework to enable this.

cs.CL↗

arXiv2024

Bi-level Trajectory Optimization on Uneven Terrains with Differentiable Wheel-Terrain Interaction Model

Navigation of wheeled vehicles on uneven terrain necessitates going beyond the 2D approaches for trajectory planning. Specifically, it is essential to incorporate the full 6dof variation of vehicle pose and its associated stability cost in the planning process. To this end, most recent works aim to learn a neural network model to predict the vehicle evolution. However, such approaches are data-intensive and fraught with generalization issues. In this paper, we present a purely model-based approach that just requires the digital elevation information of the terrain. Specifically, we express the wheel-terrain interaction and 6dof pose prediction as a non-linear least squares (NLS) problem. As a result, trajectory planning can be viewed as a bi-level optimization. The inner optimization layer predicts the pose on the terrain along a given trajectory, while the outer layer deforms the trajectory itself to reduce the stability and kinematic costs of the pose. We improve the state-of-the-art in the following respects. First, we show that our NLS based pose prediction closely matches the output from a high-fidelity physics engine. This result coupled with the fact that we can query gradients of the NLS solver, makes our pose predictor, a differentiable wheel-terrain interaction model. We further leverage this differentiability to efficiently solve the proposed bi-level trajectory optimization problem. Finally, we perform extensive experiments, and comparison with a baseline to showcase the effectiveness of our approach in obtaining smooth, stable trajectories.

cs.RO↗

arXiv2024

Who Evaluates the Evaluations? Objectively Scoring Text-to-Image Prompt Coherence Metrics with T2IScoreScore (TS2)

With advances in the quality of text-to-image (T2I) models has come interest in benchmarking their prompt faithfulness -- the semantic coherence of generated images to the prompts they were conditioned on. A variety of T2I faithfulness metrics have been proposed, leveraging advances in cross-modal embeddings and vision-language models (VLMs). However, these metrics are not rigorously compared and benchmarked, instead presented with correlation to human Likert scores over a set of easy-to-discriminate images against seemingly weak baselines. We introduce T2IScoreScore, a curated set of semantic error graphs containing a prompt and a set of increasingly erroneous images. These allow us to rigorously judge whether a given prompt faithfulness metric can correctly order images with respect to their objective error count and significantly discriminate between different error nodes, using meta-metric scores derived from established statistical tests. Surprisingly, we find that the state-of-the-art VLM-based metrics (e.g., TIFA, DSG, LLMScore, VIEScore) we tested fail to significantly outperform simple (and supposedly worse) feature-based metrics like CLIPScore, particularly on a hard subset of naturally-occurring T2I model errors. TS2 will enable the development of better T2I prompt faithfulness metrics through more rigorous comparison of their conformity to expected orderings and separations under objective criteria.

cs.CV↗

arXiv2024

GeoCoder: Solving Geometry Problems by Generating Modular Code through Vision-Language Models

Geometry problem-solving demands advanced reasoning abilities to process multimodal inputs and employ mathematical knowledge effectively. Vision-language models (VLMs) have made significant progress in various multimodal tasks. Yet, they still struggle with geometry problems and are significantly limited by their inability to perform mathematical operations not seen during pre-training, such as calculating the cosine of an arbitrary angle, and by difficulties in correctly applying relevant geometry formulas. To overcome these challenges, we present GeoCoder, which leverages modular code-finetuning to generate and execute code using a predefined geometry function library. By executing the code, we achieve accurate and deterministic calculations, contrasting the stochastic nature of autoregressive token prediction, while the function library minimizes errors in formula usage. We also propose a multimodal retrieval-augmented variant of GeoCoder, named RAG-GeoCoder, which incorporates a non-parametric memory module for retrieving functions from the geometry library, thereby reducing reliance on parametric memory. Our modular code-finetuning approach enhances the geometric reasoning capabilities of VLMs, yielding an average improvement of over 16% across various question complexities on the GeomVerse dataset compared to other finetuning methods.

cs.CL↗

arXiv2024

Losing Visual Needles in Image Haystacks: Vision Language Models are Easily Distracted in Short and Long Contexts

We present LoCoVQA, a dynamic benchmark generator for evaluating long-context extractive reasoning in vision language models (VLMs). LoCoVQA augments test examples for mathematical reasoning, VQA, and character recognition tasks with increasingly long visual contexts composed of both in-distribution and out-of-distribution distractor images. Across these tasks, a diverse set of VLMs rapidly lose performance as the visual context length grows, often exhibiting a striking logarithmic decay trend. This test assesses how well VLMs can ignore irrelevant information when answering queries -- a task that is quite easy for language models (LMs) in the text domain -- demonstrating that current state-of-the-art VLMs lack this essential capability for many long-context applications.

cs.CL↗

arXiv2024

Standing on FURM ground -- A framework for evaluating Fair, Useful, and Reliable AI Models in healthcare systems

The impact of using artificial intelligence (AI) to guide patient care or operational processes is an interplay of the AI model's output, the decision-making protocol based on that output, and the capacity of the stakeholders involved to take the necessary subsequent action. Estimating the effects of this interplay before deployment, and studying it in real time afterwards, are essential to bridge the chasm between AI model development and achievable benefit. To accomplish this, the Data Science team at Stanford Health Care has developed a Testing and Evaluation (T&E) mechanism to identify fair, useful and reliable AI models (FURM) by conducting an ethical review to identify potential value mismatches, simulations to estimate usefulness, financial projections to assess sustainability, as well as analyses to determine IT feasibility, design a deployment strategy, and recommend a prospective monitoring and evaluation plan. We report on FURM assessments done to evaluate six AI guided solutions for potential adoption, spanning clinical and operational settings, each with the potential to impact from several dozen to tens of thousands of patients each year. We describe the assessment process, summarize the six assessments, and share our framework to enable others to conduct similar assessments. Of the six solutions we assessed, two have moved into a planning and implementation phase. Our novel contributions - usefulness estimates by simulation, financial projections to quantify sustainability, and a process to do ethical assessments - as well as their underlying methods and open source tools, are available for other healthcare systems to conduct actionable evaluations of candidate AI solutions.

cs.CY↗

arXiv2024

ATPPNet: Attention based Temporal Point cloud Prediction Network

Point cloud prediction is an important yet challenging task in the field of autonomous driving. The goal is to predict future point cloud sequences that maintain object structures while accurately representing their temporal motion. These predicted point clouds help in other subsequent tasks like object trajectory estimation for collision avoidance or estimating locations with the least odometry drift. In this work, we present ATPPNet, a novel architecture that predicts future point cloud sequences given a sequence of previous time step point clouds obtained with LiDAR sensor. ATPPNet leverages Conv-LSTM along with channel-wise and spatial attention dually complemented by a 3D-CNN branch for extracting an enhanced spatio-temporal context to recover high quality fidel predictions of future point clouds. We conduct extensive experiments on publicly available datasets and report impressive performance outperforming the existing methods. We also conduct a thorough ablative study of the proposed architecture and provide an application study that highlights the potential of our model for tasks like odometry estimation.

cs.RO↗

arXiv2024

OCTO+: A Suite for Automatic Open-Vocabulary Object Placement in Mixed Reality

One key challenge in Augmented Reality is the placement of virtual content in natural locations. Most existing automated techniques can only work with a closed-vocabulary, fixed set of objects. In this paper, we introduce and evaluate several methods for automatic object placement using recent advances in open-vocabulary vision-language models. Through a multifaceted evaluation, we identify a new state-of-the-art method, OCTO+. We also introduce a benchmark for automatically evaluating the placement of virtual objects in augmented reality, alleviating the need for costly user studies. Through this, in addition to human evaluations, we find that OCTO+ places objects in a valid region over 70% of the time, outperforming other methods on a range of metrics.

cs.CV↗

arXiv2023

OCTOPUS: Open-vocabulary Content Tracking and Object Placement Using Semantic Understanding in Mixed Reality

One key challenge in augmented reality is the placement of virtual content in natural locations. Existing automated techniques are only able to work with a closed-vocabulary, fixed set of objects. In this paper, we introduce a new open-vocabulary method for object placement. Our eight-stage pipeline leverages recent advances in segmentation models, vision-language models, and LLMs to place any virtual object in any AR camera frame or scene. In a preliminary user study, we show that our method performs at least as well as human experts 57% of the time.

cs.CV↗

arXiv2023

NeuroSMPC: A Neural Network guided Sampling Based MPC for On-Road Autonomous Driving

In this paper we show an effective means of integrating data driven frameworks to sampling based optimal control to vastly reduce the compute time for easy adoption and adaptation to real time applications such as on-road autonomous driving in the presence of dynamic actors. Presented with training examples, a spatio-temporal CNN learns to predict the optimal mean control over a finite horizon that precludes further resampling, an iterative process that makes sampling based optimal control formulations difficult to adopt in real time settings. Generating control samples around the network-predicted optimal mean retains the advantage of sample diversity while enabling real time rollout of trajectories that avoids multiple dynamic obstacles in an on-road navigation setting. Further the 3D CNN architecture implicitly learns the future trajectories of the dynamic agents in the scene resulting in successful collision free navigation despite no explicit future trajectory prediction. We show performance gain over multiple baselines in a number of on-road scenes through closed loop simulations in CARLA. We also showcase the real world applicability of our system by running it on our custom Autonomous Driving Platform (AutoDP).

cs.RO↗

arXiv2023

Hilbert Space Embedding-based Trajectory Optimization for Multi-Modal Uncertain Obstacle Trajectory Prediction

Safe autonomous driving critically depends on how well the ego-vehicle can predict the trajectories of neighboring vehicles. To this end, several trajectory prediction algorithms have been presented in the existing literature. Many of these approaches output a multi-modal distribution of obstacle trajectories instead of a single deterministic prediction to account for the underlying uncertainty. However, existing planners cannot handle the multi-modality based on just sample-level information of the predictions. With this motivation, this paper proposes a trajectory optimizer that can leverage the distributional aspects of the prediction in a computationally tractable and sample-efficient manner. Our optimizer can work with arbitrarily complex distributions and thus can be used with output distribution represented as a deep neural network. The core of our approach is built on embedding distribution in Reproducing Kernel Hilbert Space (RKHS), which we leverage in two ways. First, we propose an RKHS embedding approach to select probable samples from the obstacle trajectory distribution. Second, we rephrase chance-constrained optimization as distribution matching in RKHS and propose a novel sampling-based optimizer for its solution. We validate our approach with hand-crafted and neural network-based predictors trained on real-world datasets and show improvement over the existing stochastic optimization approaches in safety metrics.

cs.RO↗

arXiv2023

TwiRGCN: Temporally Weighted Graph Convolution for Question Answering over Temporal Knowledge Graphs

Recent years have witnessed much interest in temporal reasoning over knowledge graphs (KG) for complex question answering (QA), but there remains a substantial gap in human capabilities. We explore how to generalize relational graph convolutional networks (RGCN) for temporal KGQA. Specifically, we propose a novel, intuitive and interpretable scheme to modulate the messages passed through a KG edge during convolution, based on the relevance of its associated time period to the question. We also introduce a gating device to predict if the answer to a complex temporal question is likely to be a KG entity or time and use this prediction to guide our scoring mechanism. We evaluate the resulting system, which we call TwiRGCN, on TimeQuestions, a recently released, challenging dataset for multi-hop complex temporal QA. We show that TwiRGCN significantly outperforms state-of-the-art systems on this dataset across diverse question types. Notably, TwiRGCN improves accuracy by 9--10 percentage points for the most difficult ordinal and implicit question types.

cs.CL↗

arXiv2023

SAppKG: Mobile App Recommendation Using Knowledge Graph and Side Information-A Secure Framework

Due to the rapid development of technology and the widespread usage of smartphones, the number of mobile applications is exponentially growing. Finding a suitable collection of apps that aligns with users needs and preferences can be challenging. However, mobile app recommender systems have emerged as a helpful tool in simplifying this process. But there is a drawback to employing app recommender systems. These systems need access to user data, which is a serious security violation. While users seek accurate opinions, they do not want to compromise their privacy in the process. We address this issue by developing SAppKG, an end-to-end user privacy-preserving knowledge graph architecture for mobile app recommendation based on knowledge graph models such as SAppKG-S and SAppKG-D, that utilized the interaction data and side information of app attributes. We tested the proposed model on real-world data from the Google Play app store, using precision, recall, mean absolute precision, and mean reciprocal rank. We found that the proposed model improved results on all four metrics. We also compared the proposed model to baseline models and found that it outperformed them on all four metrics.

cs.SI↗

arXiv2023

Avalanche Prediction and Dynamics using Temperature Variance , Grain Size Variance and Flow Regimes

We investigate the effects of temperature variance, grain size variation, flow regimes, and the use of Support Vector Machines (SVMs) in avalanche studies. The temperature variance experiments involved ice single crystals and polycrystals, revealing that the scale-free pattern of avalanche sizes remains consistent regardless of temperature. The dynamics of dislocations in polycrystals were found to be independent of stress level and temperature. The Material Point Method (MPM) was used to explore snow avalanche behavior and identify flow regimes. The MPM accurately represented various flow patterns of snow avalanches, although challenges remained in capturing powder clouds. SVMs were employed for avalanche forecasting, using meteorological and snowpack variables as input features. The selected features provided insights into snowfall characteristics, snow accumulation, rain interaction, snowdrift patterns, cloud dynamics, snowpack mechanics, and temperature distribution within the snowpack. The findings contribute to a better understanding of avalanche dynamics and offer potential improvements in avalanche trend predictions.

physics.geo-ph↗

arXiv2023

Towards Carrollian quantization: renormalization of Carrollian electrodynamics

Field-theoretic description of Carrollian theories has largely remained classical so far. In this paper, we attempt to study the renormalization of Carrollian gauge field theories via path integral techniques. The case of Carrollian electrodynamics minimally coupled to a massive Carrollian scalar is considered. We report potential problems such as IR divergences and mass shell singularity cropping up at the first order in the perturbation. Perhaps, the most important result that we report is how conventional arguments for gauge independence for mass and coupling are invalidated for a gauge theory in a Carrollian setting. As of now, the renormalization of Carrollian gauge field theories seems to suffer from unphysical ramifications. Possible cures to resolve these issues are suggested.

hep-th↗

arXiv2023

Galilean fermions: Classical and quantum aspects

We study the classical and quantum "properties" of Galilean fermions in 3+1 dimensions. We have taken the case of massless Galilean fermions minimally coupled to the scalar field. At the classical level, the Lagrangian is obtained by null reducing the relativistic theory in one higher dimension. The resulting theory is found to be invariant under infinite Galilean conformal symmetries. Using Noether's procedure, we construct the corresponding infinite conserved charges. Path integral techniques are then employed to probe the quantum "properties" of the theory. The theory is found to be renormalizable. A novel feature of the theory is the emergence of mass scale at the first order of quantum correction. The conformal symmetry of the theory breaks at the quantum level. We confirm this by constructing the beta function of the theory.

hep-th↗