Neue Publikationen in Ziel-Venues – 2026-11-02
Neue Werke: 5 · davon mit DE-Affiliation: 5 Venues: 60 · abgefragt ab publication_date 2026-10-03 · letzter Lauf: 2026-11-02T12:03:54
Design, Automation and Test in Europe (DATE) — 1 neu · conference
TrackScorer: Skyrmion Logic-in-Memory Accelerator for Tree-Based Ranking Models ⚑ DE
- DOI: 10.23919/date64628.2025.10992934
- Metadaten: Erschienen: 2025-03-31 · S. 1-7 · OpenAlex seit 2025-05-22
- DE-Institutionen: Lamarr Institute for Machine Learning and Artificial Intelligence
- Autoren:
- Elijah Seth Cishugi — University of Twente (NL)
- Sebastian Buschjäger — Lamarr Institute for Machine Learning and Artificial Intelligence (DE)
- Martijn Noorlander — University of Twente (NL)
- Marco Luigi Ottavi — University of Twente (NL)
- Kuan-Hsun Chen — University of Twente (NL)
- Topics: Parallel Computing and Optimization Techniques; Advanced Memory and Neural Computing; Magnetic properties of thin films
- Keywords: racetrack memory; document ranking; quadratic assignment problem; inference acceleration; leakage power reduction
- Abstract: Racetrack memories (RTMs) have been shown to have lower leakage power and higher density compared to traditional DRAM/SRAM technologies. However, their efficiency is often hindered by the need to shift the targeted data to access ports for read and write operations. Suitable mapping approaches are therefore essential to unleash their potential. In this work, we explore the mapping of the popular tree-based document ranking algorithm, Quickscorer, onto Skyrmion-based racetrack memories (SK-RTMs). Our approach leverages a Logic-in-Memory (LiM) accelerator, specifically designed to execute simple logic operations directly within SK-RTMs, enabling an efficient mapping of Quickscorer by exploiting its bitvector representation and inter-leaved traversal scheme of tree structures through bitwise logical operations. We present several mapping strategies, including one based on a quadratic assignment problem (QAP) optimization algorithm for optimal data placement of Quickscorer onto the racetracks. Our results demonstrate a significant reduction in read and write operations and, in certain cases, a decrease in the time spent shifting data during Quickscorer inference.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) — 3 neu · conference
Estimating Respiratory Effort from Nocturnal Breathing Sounds for Obstructive Sleep Apnoea Screening ⚑ DE
- DOI: 10.1109/icassp55912.2026.11460863
- Metadaten: Erschienen: 2026-04-21 · S. 14962-14966 · OpenAlex seit 2025-10-10
- DE-Institutionen: Max Planck School Matter to Life
- Autoren:
- Xiaolei Xu — University of Sheffield (GB)
- Chaoyue Niu — University of Sheffield (GB)
- Guy J. Brown — University of Sheffield (GB)
- Hector E. Romero — Max Planck School Matter to Life (DE)
- Ning Ma — University of Sheffield (GB)
- Topics: Obstructive Sleep Apnea Research; Non-Invasive Vital Sign Monitoring; Phonocardiography and Auscultation Techniques
- Keywords: obstructive sleep apnea; smartphone audio; apnea-hypopnea index; home sleep monitoring
- Abstract: Obstructive sleep apnoea (OSA) is a prevalent condition with significant health consequences, yet many patients remain undiagnosed due to the complexity and cost of overnight polysomnography. Acoustic-based screening provides a scalable alternative, yet performance is limited by environmental noise and the lack of physiological context. Respiratory effort is a key signal used in clinical scoring of OSA events, but current approaches require additional contact sensors that reduce scalability and patient comfort. This paper presents the first study to estimate respiratory effort directly from nocturnal audio, enabling physiological context to be recovered from sound alone. We propose a latent-space fusion framework that integrates the estimated effort embeddings with acoustic features for OSA detection. Using a dataset of 157 nights from 103 participants recorded in home environments, our respiratory effort estimator achieves a concordance correlation coefficient of 0.48, capturing meaningful respiratory dynamics. Fusing effort and audio improves sensitivity and AUC over audio-only baselines, especially at low apnoea–hypopnoea index thresholds. The proposed approach requires only smartphone audio at test time, which enables sensorfree, scalable, and longitudinal OSA monitoring.
Are Modern Speech Enhancement Systems Vulnerable to Adversarial Attacks? ⚑ DE
- DOI: 10.1109/icassp55912.2026.11464211
- Metadaten: Erschienen: 2026-04-21 · S. 15762-15766 · OpenAlex seit 2026-04-22
- DE-Institutionen: Universität Hamburg; Helmholtz Center for Information Security
- Autoren:
- Rostislav Makarov — Universität Hamburg (DE)
- Lea Schönherr — Helmholtz Center for Information Security (DE)
- Timo Gerkmann — Universität Hamburg (DE)
- Topics: Adversarial Robustness in Machine Learning; Speech and Audio Processing; Speech Recognition and Synthesis
- Keywords: speech enhancement; adversarial attacks; adversarial perturbations; semantic manipulation; diffusion models; robustness; psychoacoustic masking; machine learning security
- Abstract: Machine learning approaches for speech enhancement are becoming increasingly expressive, enabling ever more powerful modifications of input signals. In this paper, we demonstrate that this expressiveness introduces a vulnerability: advanced speech enhancement models can be susceptible to adversarial attacks. Specifically, we show that adversarial noise, carefully crafted and psychoacoustically masked by the original input, can be injected such that the enhanced speech output conveys an entirely different semantic meaning. We experimentally verify that contemporary predictive speech enhancement models can indeed be manipulated in this way. Furthermore, we highlight that diffusion models with stochastic samplers exhibit inherent robustness to such adversarial attacks by design.
Addressing Pilot Contamination in Channel Estimation with Variational Autoencoders ⚑ DE
- DOI: 10.1109/icassp49660.2025.10888089
- Metadaten: Erschienen: 2025-03-12 · S. 1-5 · OpenAlex seit 2025-03-13
- DE-Institutionen: Technical University of Munich
- Autoren:
- Amar Kasibovic — Technical University of Munich (DE)
- Benedikt Fesl — Technical University of Munich (DE)
- Michael Baur — Technical University of Munich (DE)
- Wolfgang Utschick — Technical University of Munich (DE)
- Topics: Advanced MIMO Systems Optimization; Advanced Wireless Communication Techniques; Millimeter-Wave Propagation and Modeling
- Keywords: pilot contamination; channel estimation; massive MIMO; variational autoencoder; inter-cell interference; pilot reuse; statistical channel knowledge
- Abstract: Pilot contamination (PC) is a well-known problem that affects massive multiple-input multiple-output (MIMO) systems. When frequency and pilots are reused between different cells, PC constitutes one of the main bottlenecks of the system's performance. In this paper, we propose a method based on the variational autoencoder (VAE), capable of reducing the impact of PC-related interference during channel estimation (CE). We obtain the first and second-order statistics of the conditionally Gaussian (CG) channels for both the user equipments (UEs) in a cell of interest and those in interfering cells, and we then use these moments to compute conditional linear minimum mean square error estimates. We show that the proposed estimator is capable of exploiting the interferers’ additional statistical knowledge, outperforming other classical approaches. Moreover, we highlight how the achievable performance is tied to the chosen setup, making the setup selection crucial in the study of multi-cell CE.
Advanced Optical Materials (AOM) — 1 neu · materials
Strong Enhancement of Er 3+ Emission at Room Temperature in Si 3 N 4 Metasurfaces ⚑ DE
- DOI: 10.1002/adom.71998
- Metadaten: Erschienen: 2026-10-30 · OpenAlex seit 2026-05-06
- DE-Institutionen: Friedrich Schiller University Jena
- Autoren:
- Fengkai Wei — Friedrich Schiller University Jena (DE)
- Xinru Ji — École Polytechnique Fédérale de Lausanne (CH)
- Tobias Jan Kippenberg — École Polytechnique Fédérale de Lausanne (CH)
- Duk‐Yong Choi — Australian National University (AU)
- Carsten Ronning — Friedrich Schiller University Jena (DE)
- Topics: Metamaterials and Metasurfaces Applications; Photonic and Optical Devices; Luminescence Properties of Advanced Materials
- Keywords: metasurfaces; nanocylinder arrays; Mie resonance; local density of optical states; Purcell effect; ion implantation depth; luminescence lifetime; telecom wavelengths
- Abstract: ABSTRACT We report room‐temperature photoluminescence (PL) enhancement at from ion‐implanted Er 3+ ions in Mie‐resonant silicon nitride metasurfaces. Following annealing at , nanocylinders with a radius of exhibit an 18.1‐fold detected PL enhancement relative to a planar film. The measured lifetime decreases from to , corresponding to approximately 9.4‐fold shortening. These observations and electromagnetic simulations are consistent with modified spontaneous‐emission dynamics, although excitation, non‐radiative relaxation, and collection also influence the detected intensity. Annealing increases PL and narrows the radius‐dependent enhancement peak: Gaussian fits give geometric full widths at half maximum of approximately before annealing and after annealing. These widths describe the PL enhancement peak; linking them to the optical resonance requires an independent linewidth measurement. Across implantation energies of , depth and total implanted dose vary together, so the increased PL tracks the larger implanted Er population. This CMOS‐compatible platform combines enhanced telecom‐band Er emission with scalable fabrication and room‐temperature operation, offering potential for compact, integrated light sources.