Shocker Open Access Repository
Featured Items
Communities in SOAR
Select a community to browse its collections.
Recent Submissions
Item Metadata only Phonetic Analysis of Real and Synthetic Speech Using HuBERT Embeddings: Perspectives for Deepfake Detection(Institute of Electrical and Electronics Engineers Inc., 2026-01-28)The growing sophistication of speech generated by Artificial Intelligence (AI) has introduced new challenges in audio deepfake detection. Text-to-speech (TTS) and voice conversion (VC) technologies can now produce convincing synthetic speech with high quality and intelligibility. This poses a serious threat to voice biometric security systems, such as automatic speaker recognition. It also increases the risks associated to the spread of spoken disinformation, where synthetic voices can be used to disseminate malicious content. In this study, we conduct an analysis of real and synthetic speech at phonetic and word levels. For that, a parallel dataset comprising real and synthetic speech signals were developed based on a subset of the LibriSpeech ASR corpus. Synthetic speech samples were generated using two TTS and one VC systems: Coqui TTS, VITS TTS, and StarGANv2 VC. We adopted HuBERT, a self-supervised speech model, to extract speech embeddings. The motivation for using this model stems from its ability to recognize sound units corresponding to the so-called pseudo phonemes. Our analysis is based on the KL divergence (KLD) between the distributions of synthetic and real phonemes, which allowed us to rank synthetic phonemes based on their alignment with their real counterpart. We also trained several classifiers per phoneme to distinguish between real and synthetic samples. We then compute the correlations between KLD and accuracies per phoneme. Besides showing a list of phonemes that are more discriminative, our findings suggest that vowels correlate better with the classifiers' performance, suggesting that the KLD can be an indicator of the most distinguishable phonemes for deepfake detection. © 2025 IEEE.Item Open Access Reinforcing Edge-DASH: Deep learning for multi-objective streaming optimization(Institute of Electrical and Electronics Engineers Inc., 2026-03-16)With the growing demand for multimedia services, Dynamic Adaptive Streaming over HTTP (DASH) has become a key solution for delivering high-quality video content. In this work, we consider an Edge-DASH scenario and formulate a joint optimization problem that involves four critical aspects: bitrate allocation, user-to-server assignment, caching, and bandwidth allocation. Due to the complexity of the joint problem, we decompose it into sub-problems and address them separately. To solve the resulting sub-problems, we employ deep reinforcement learning, specifically the Deep Deterministic Policy Gradient (DDPG) method, for three of them, and develop a heuristic solution for the fourth. Simulation results demonstrate that our approach enhances performance across multiple metrics, including improved video delivery, reduced buffer underflow and overflow, and more efficient caching, which collectively enable greater utilization of edge resources for streaming. Moreover, we evaluated inference latency across edge and cloud hardware, confirming sub- to few-millisecond performance suitable for real-time deployment. This showcases the benefits of combining learning-based and heuristic techniques to meet the growing demand for adaptive video streaming in edge computing environments. © 2002-2012 IEEE.Item Metadata only Time-series pattern mining in SWIFT logs using AI-driven sequential embeddings built in python(CRC Press, 2026)Analyzing financial communication streams like SWIFT logs can provide insight into potential anomalies, compliance issues, or operational inefficiencies. Traditional time-series analysis techniques often overlook the complex structures and latent semantics within interdependent message flows. Here, we describe our work on designing a Python architecture for mining temporal patterns in SWIFT logs employing transformer-based sequential embeddings. Our method divides the message flows into time-windowed sequences, transforms the raw SWIFT fields into contextual embeddings, then applies supervised anomaly detection and unsupervised clustering to high-risk behavioural motif extraction. We tested our model on a dataset containing 3.1 million real-life SWIFT messages over an 18-month period. The hybrid transformer model proposed in this work surpasses the traditional LSTM autoencoders and baseline token classifiers, achieving a 94% F1 score in anomaly classification. The embedding clusters and temporal heatmaps visualizations show the model’s known compliance flags alongside previously hidden irregular patterns. Moreover, the system achieves real-time performance constraints, processing message batches with very low latency while adapting to changes in the streaming data. These results are promising for leveraging powerful sequence models in discovering intricate patterns in financial transaction logs and indicate potential advancements in AI-driven predictive compliance and risk analytics within SWIFT institutional frameworks. © 2026 selection and editorial matter, Dr. Raja M., Dr. Satya Subrahmanyam, Dr. R. Raja Subramanian and Dr. J. Karthikeyan; individual chapters, the contributors.Item Metadata only Structurally driven, reversible topological phase transition in a distorted square net material(American Physical Society, 2026-03-17)Topological materials hold immense promise for exhibiting exotic quantum phenomena, yet achieving controllable topological phase transitions remains challenging. Here, we demonstrate a structurally driven, reversible topological phase transition in the distorted square net material GdPS, induced via in situ potassium dosing. Using angle-resolved photoemission spectroscopy and first principles calculations, we demonstrate a cascade of topological phases in the subsurface P layer: from a large, topologically trivial band gap to a gapless Dirac cone state with a 2 eV dispersion, and finally to a two-dimensional topological insulator as inferred from theory. This evolution is driven by subtle structural distortions in the first P layer caused by potassium adsorption, which in turn contribute to the band gap closure and topological phase transition. Furthermore, the ability to manipulate the topology of a subsurface layer in GdPS offers a unique route for exploring and controlling topological states in bulk materials. © 2026 American Physical Society.Item Metadata only Chalcogen doping effect on the insulator-to-metal transition in GdPS(Elsevier Ltd, 2026-03-17)Topological semimetals offer a rich platform for exploring massless fermion physics and realizing exotic properties with potential technological applications. GdPS, a magnetic semiconductor derived from the nodal-line semimetal ZrSiS family, exhibits a field-induced insulator-to-metal transition driven by exchange splitting. This transition is accompanied by an unusual, isotropic, and gigantic negative magnetoresistance, attributed to negligible magnetic anisotropy resulting from the weak spin-orbit coupling of half-filled Gd³ ⁺ 4 f orbitals and light S atoms. In this work, we investigate Se substitution, which is expected to enhance spin-orbit coupling. Indeed, we observe slightly increased magnetic anisotropy in magnetotransport. Moreover, Se substitution suppresses the field-induced insulator-to-metal transition, likely due to an enlarged band gap that demands a higher exchange splitting to close. These findings provide deeper insights into the interplay between spin-orbit coupling, magnetic anisotropy, and transport behavior in GdPS, offering guidance for future materials design for desired functionalities. © 2026 The Authors