Clay weekly context brief for the Systems category (ISO week 2026-W43). Clay tracks publications from the Systems feed list. Below are recent items from this category, each with its source and a short description of what the publication covers when one is available in the source feed. Recent publications: 1. LightHybridNet: A Lightweight Hybrid Deep Learning Framework for Efficient Multi-Label Ophthalmic Disease Classification in Imbalanced Fundus Images Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.20671 Eye diseases are a major global cause of blindness and vision impairment, highlighting the need for accurate and early diagnosis to prevent further deterioration. 2. Evolution Toward Sensing-Enabled RAN: Practical Bistatic Sensing via Uplink Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.20923 This paper considers uplink channel state information (CSI)-based sensing as an opportunistic approach to enabling sensing in cellular networks, with minimum required modifications to the infrastructure. 3. Continued fractions and passive ladder networks for acoustic impedances: the baffled piston and the horn Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.20371 Lumped-element networks give transparent, passive models of acoustic impedances whose low- and high-frequency limits can be read from the circuit, but a systematic, arbitrarily accurate construction for non-rational impedances is lacking. 4. Separating mixtures of single-fiber orientation distributions for increased sensitivity to crossing white matter fibers Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.20974 Intro: Diffusion MRI allows for in vivo probing of brain tissue architecture via micron-level diffusion aggregated at the millimeter scale, but crossing white matter fibers decrease the sensitivity of downstream tasks such as microstructural modeling and tractography. 5. Resilient and Efficient Aerial Wireless Sensor Networks Using Pinching-Antenna Systems Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.20924 This paper proposes a pinching antenna system to enhance the connectivity and efficiency in aerial wireless networks. 6. Learning Visual Speech Representations via Cross-Modal Distillation and Joint Face-Lip Modeling Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.19559 Lipreading remains challenging due to the inherent ambiguity of visual speech cues and their sensitivity to head pose and video resolution. 7. Evidence Field for Histopathology: Cross-Centre Evaluation and Replayable Spatial Decisions Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.21115 Histopathology classifiers must transfer across acquisition sites while exposing evidence that can be checked independently. 8. Range Ambiguity Identification and Elimination by Leveraging Temporal Broadening Effect for THz Sensing Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.20928 High pulse repetition frequency (PRF) operation improves Doppler unambiguous sensing but reduces the maximum unambiguous range of a terahertz (THz) pulsed radar. 9. Designing Textual Interfaces To Enhance Llm Reasoning On Audio-Visual Question Answering Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.20085 Automatic prompt optimization (APO) has shown strong effectiveness for text-based LLMs and LLM systems, but its application to audio-visual understanding systems requires suitable textual interfaces through which multimodal reasoning can be exposed to text-based optimizers. 10. Tokenized Joint Source-Network Coding for Loss-Resilient Video Streaming Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.21166 For decades, conventional video streaming systems have relied on packet-level forward error correction (FEC) codes and retransmissions to ensure reliable video delivery. 11. How large an effect would we have seen? Positive controls, detection floors and dataset pitfalls in scalp-EEG speech decoding Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.20942 Objective. 12. Repairing Speech Access After Text Updates with Frozen Affine Fields Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.20253 Speech-capable language models inherit most of their knowledge from text-rich language backbones, and continued model maintenance is likewise dominated by text updates. 13. Monotone Piecewise-Linear Harmonization of Multitemporal Multispectral Surface Reflectance With Unsupervised Deep Multiview Pseudoinvariant Features Learning Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.21183 Surface reflectance products derived from optical satellite may exhibit systematic inconsistencies across acquisition dates, because atmospheric correction is generally performed independently for each image. 14. Cross-subject decoding of imagined speech from scalp EEG: a leakage-controlled evaluation on three public corpora Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.20954 Objective. 15. Energy-Based Models For Speech Synthesis Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2310.12765 Recently there has been a lot of interest in non-autoregressive (non-AR) models for speech synthesis, such as FastSpeech 2 and diffusion models. 16. Bayesian Joint Reconstruction and Uncertainty Quantification for Limited-Data Multi-Contrast X-Ray CT Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.21406 Multi-contrast X-ray computed tomography (CT) combines attenuation contrast with phase and dark-field contrast, which are sensitive to refraction and ultra-small-angle scattering, respectively. 17. Antenna Placement for Movable Antenna-Enabled Near-Field Target Localization Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21191 Movable antennas (MAs) have recently emerged as a promising technology to enhance the wireless sensing performance by exploiting additional spatial degrees of freedom. 18. Score-Based Training for Energy-Based TTS Models Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2505.13771 Noise contrastive estimation (NCE) is a popular method for training energy-based models (EBM) with intractable normalisation terms. 19. Learning Variable Metrics for Convergent Proximal Splitting Algorithms Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.22204 We propose learned variable metric (LVM), a framework that learns metrics of proximal splitting algorithms adapted to each problem instance, coordinate, and iteration with a convergence guarantee. 20. Micro-Doppler Sensing for UAV Flight-State Estimation in ISAC Networks Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21225 Integrated sensing and communication (ISAC) provides a promising foundation for low-altitude unmanned aerial vehicle (UAV) monitoring. 21. Geometric Alignment without Functional Equivalence: A Layer-wise Analysis of the Speech-Text Modality Gap Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2603.01502 End-to-end speech large language models (LLMs) often answer the same question less accurately from speech than from text, even when the spoken content matches the text prompt. 22. Denoising Textures is a Different Ballgame Altogether! Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.22258 The human brain effortlessly deciphers colors and textures from visual stimuli and reconstructs the world around us, yet preserving such details remains challenging in image denoising. 23. Finite-History Outage Certification for Partially Observed Fluid Antenna Receivers Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21252 Fluid antenna receivers can select a service position from partial port observations, but finite historical data leave uncertainty in the covariance used for that decision. 24. SonicWorld: Source-Centric Latent Dynamics for Spatial Audio Reasoning Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.17131 Dynamic spatial audio reasoning requires a model to associate changing acoustic observations with the correct source and recover its spatial state at a queried time or interval, a challenge intensified by overlapping speech and source motion. 25. Data-driven Smooth-to-hard Spectral Regularization for General Linear Inverse Problems Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2610.21664 The logistic smooth-to-hard shrinker is the proximal map of an explicit saturating spectral penalty for general linear inverse problems. 26. Receiver-Specific Interventional Valuation of Semantic Atoms for Budget-Constrained Multi-Agent Communication Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21504 Cooperative multi-agent semantic communication requires deciding not only how a transmitter encodes its observation, but also which resulting semantic content should be delivered to each receiver under a limited link budget. 27. FAM-gate: field-aware metadata attention with prior-constrained vector gating for audio classification Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.14673 User-generated audio collections frequently contain textual metadata such as titles, tags, and descriptions. 28. PanoDiff-SR: Synthesizing Dental Panoramic Radiographs using Diffusion and Super-resolution Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2507.09227 Interest in generating realistic synthetic medical images has grown in recent years. 29. Wavefront Engineering for Physical Layer Security in Near-Field Wireless Communications Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21521 With wireless transmission exposing information to potential interception by unintended receivers, physical layer security (PLS) offers a promising approach to securing wireless communications. 30. Cross-Lingual TTS for Indian Languages: A Benchmark and Inference-Time Duration Estimation Strategies Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.14826 Recent alignment-free Text-to-Speech (TTS) models achieve high synthesis quality by removing explicit alignment modules and simplifying training. 31. Absorption and Phase-Contrast Microtomography Using Direct X-Ray Detection With Commercial Off-The-Shelf CMOS Sensors Source: eess.IV (Image and Video Processing) Link: https://arxiv.org/abs/2605.29808 This work presents high-resolution X-ray microtomography based on commercial off-the-shelf (COTS) CMOS image sensors as direct detectors. 32. From Interference to Diversity: Mode Coloring and Two-Dimensional Modal-Time Rake Combining for Multipath OAM Communications Source: eess.SP (Signal Processing) Link: https://arxiv.org/abs/2610.21602 Orbital angular momentum (OAM) beams provide a spatial-multiplexing dimension for wireless communications, but multipath scattering broadens their mode spectrum, so that energy leaks across modes as inter-mode interference (IMI), which conventional OAM demultiplexing discards. 33. SemiCodec: Bridging Speech Tokenization and High-fidelity Reconstruction with Semi-discrete Representations Source: eess.AS (Audio and Speech Processing) Link: https://arxiv.org/abs/2610.14834 Speech representation learning requires a discrete modeling interface while preserving high waveform fidelity. Sources in this brief: eess.AS (Audio and Speech Processing); eess.IV (Image and Video Processing); eess.SP (Signal Processing). Selected 33 of 441 available items for this weekly brief.