Pavan Reddy

Sharan Mourya, Pavan Reddy, SaiDhiraj Amuru et al.

In the realm of 5G communication systems, the accuracy of Channel State Information (CSI) prediction is vital for optimizing performance. This letter introduces a pioneering approach: the Spectral-Temporal Graph Neural Network (STEM GNN), which fuses spatial relationships and temporal dynamics of the wireless channel using the Graph Fourier Transform. We compare the STEM GNN approach with conventional Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) models for CSI prediction. Our findings reveal a significant enhancement in overall communication system performance through STEM GNNs. For instance, in one scenario, STEM GNN achieves a sum rate of 5.009 bps/Hz which is $11.9\%$ higher than that of LSTM and $35\%$ higher than that of RNN. The spectral-temporal analysis capabilities of STEM GNNs capture intricate patterns often overlooked by traditional models, offering improvements in beamforming, interference mitigation, and ultra-reliable low-latency communication (URLLC).

Poojan Vanani, Darsh Patel, Danyal Khorami et al.

Motion capture remains costly and complex to deploy, limiting use outside specialized laboratories. We present Mesquite, an open-source, low-cost inertial motion-capture system that combines a body-worn network of 15 IMU sensor nodes with a hip-worn Android smartphone for position tracking. A low-power wireless link streams quaternion orientations to a central USB dongle and a browser-based application for real-time visualization and recording. Built on modern web technologies -- WebGL for rendering, WebXR for SLAM, WebSerial and WebSockets for device and network I/O, and Progressive Web Apps for packaging -- the system runs cross-platform entirely in the browser. In benchmarks against a commercial optical system, Mesquite achieves mean joint-angle error of 2-5 degrees while operating at approximately 5% of the cost. The system sustains 30 frames per second with end-to-end latency under 15ms and a packet delivery rate of at least 99.7% in standard indoor environments. By leveraging IoT principles, edge processing, and a web-native stack, Mesquite lowers the barrier to motion capture for applications in entertainment, biomechanics, healthcare monitoring, human-computer interaction, and virtual reality. We release hardware designs, firmware, and software under an open-source license (GNU GPL).

Aditya Singh, Pavan Reddy

Anomaly detection, a critical facet in data analysis, involves identifying patterns that deviate from expected behavior. This research addresses the complexities inherent in anomaly detection, exploring challenges and adapting to sophisticated malicious activities. With applications spanning cybersecurity, healthcare, finance, and surveillance, anomalies often signify critical information or potential threats. Inspired by the success of Anomaly Generative Adversarial Network (AnoGAN) in image domains, our research extends its principles to tabular data. Our contributions include adapting AnoGAN's principles to a new domain and promising advancements in detecting previously undetectable anomalies. This paper delves into the multifaceted nature of anomaly detection, considering the dynamic evolution of normal behavior, context-dependent anomaly definitions, and data-related challenges like noise and imbalances.

Pavan Reddy, Aditya Sanjay Gujral

Adversarial attacks in machine learning traditionally focus on global perturbations to input data, yet the potential of localized adversarial noise remains underexplored. This study systematically evaluates localized adversarial attacks across widely-used methods, including FGSM, PGD, and C&W, to quantify their effectiveness, imperceptibility, and computational efficiency. By introducing a binary mask to constrain noise to specific regions, localized attacks achieve significantly lower mean pixel perturbations, higher Peak Signal-to-Noise Ratios (PSNR), and improved Structural Similarity Index (SSIM) compared to global attacks. However, these benefits come at the cost of increased computational effort and a modest reduction in Attack Success Rate (ASR). Our results highlight that iterative methods, such as PGD and C&W, are more robust to localization constraints than single-step methods like FGSM, maintaining higher ASR and imperceptibility metrics. This work provides a comprehensive analysis of localized adversarial attacks, offering practical insights for advancing attack strategies and designing robust defensive systems.

Pavan Reddy, Nithin Reddy

In 2023, the National Eating Disorders Association's (NEDA) chatbot Tessa was suspended after providing harmful weight-loss advice to vulnerable users-an avoidable failure that underscores the risks of unsafe AI in healthcare contexts. This paper examines Tessa as a case study in absent safety engineering and demonstrates how a lightweight, modular safeguard could have prevented the incident. We propose a hybrid safety middleware that combines deterministic lexical gates with an in-line large language model (LLM) policy filter, enforcing fail-closed verdicts and escalation pathways within a single model call. Using synthetic evaluations, we show that this design achieves perfect interception of unsafe prompts at baseline cost and latency, outperforming traditional multi-stage pipelines. Beyond technical remedies, we map Tessa's failure patterns to established frameworks (OWASP LLM Top10, NIST SP 800-53), connecting practical safeguards to actionable governance controls. The results highlight that robust, auditable safety in health-adjacent AI does not require heavyweight infrastructure: explicit, testable checks at the last mile are sufficient to prevent "another Tessa", while governance and escalation ensure sustainability in real-world deployment.

Pavan Reddy, Aditya Sanjay Gujral

Large language model (LLM) assistants are increasingly integrated into enterprise workflows, raising new security concerns as they bridge internal and external data sources. This paper presents an in-depth case study of EchoLeak (CVE-2025-32711), a zero-click prompt injection vulnerability in Microsoft 365 Copilot that enabled remote, unauthenticated data exfiltration via a single crafted email. By chaining multiple bypasses-evading Microsofts XPIA (Cross Prompt Injection Attempt) classifier, circumventing link redaction with reference-style Markdown, exploiting auto-fetched images, and abusing a Microsoft Teams proxy allowed by the content security policy-EchoLeak achieved full privilege escalation across LLM trust boundaries without user interaction. We analyze why existing defenses failed, and outline a set of engineering mitigations including prompt partitioning, enhanced input/output filtering, provenance-based access control, and strict content security policies. Beyond the specific exploit, we derive generalizable lessons for building secure AI copilots, emphasizing the principle of least privilege, defense-in-depth architectures, and continuous adversarial testing. Our findings establish prompt injection as a practical, high-severity vulnerability class in production AI systems and provide a blueprint for defending against future AI-native threats.

Sharan Mourya, Pavan Reddy, SaiDhiraj Amuru et al.

Recently, deep neural networks have emerged as a solution to solve NP-hard wireless resource allocation problems in real-time. However, multi-layer perceptron (MLP) and convolutional neural network (CNN) structures, which are inherited from image processing tasks, are not optimized for wireless network problems. As network size increases, these methods get harder to train and generalize. User pairing is one such essential NP-hard optimization problem in wireless communication systems that entails selecting users to be scheduled together while minimizing interference and maximizing throughput. In this paper, we propose an unsupervised graph neural network (GNN) approach to efficiently solve the user pairing problem. Our proposed method utilizes the Erdos goes neural pipeline to significantly outperform other scheduling methods such as k-means and semi-orthogonal user scheduling (SUS). At 20 dB SNR, our proposed approach achieves a 49% better sum rate than k-means and a staggering 95% better sum rate than SUS while consuming minimal time and resources. The scalability of the proposed method is also explored as our model can handle dynamic changes in network size without experiencing a substantial decrease in performance. Moreover, our model can accomplish this without being explicitly trained for larger or smaller networks facilitating a dynamic functionality that cannot be achieved using CNNs or MLPs.