Clay weekly context brief for the Systems category (ISO week 2026-W34).

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. CASA: Content-Acoustic Speaking Assessment with Speech Encoder and Large Language Model
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.13101
   Research on automatic speaking assessment (ASA) has increasingly adopted multimodal speech large language models to assess learners' speaking performance.
2. CW-BASS v2: Saturation-Aware Pseudo-Label Selection for Semi-Supervised Segmentation under Foundation-Model Teachers
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2608.12773
   Semi-supervised semantic segmentation has long turned on one question, which pseudo-labels to trust, and a generation of selection rules, dynamic thresholds, per-class curricula, soft confidence weights, answered it for the noisy, under-confident ResNet teachers of their day.
3. Feasibility and Convex Design of Probe-Position Matching in a Scanning X-Band Radar Array
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12602
   The probe position of a microstrip element is normally chosen from the isolated-element input resistance.
4. Motor, Cognitive, or Corpus? What Survives Cross-Lingual Transfer in Speech-Based Parkinsons Disease Detection
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.13425
   Self-supervised learning (SSL) speech representations achieve strong performance for Parkinson's disease (PD) detection within individual corpora.
5. CardioState-JEPA: Delay-Aware Cross-Modal Learning of a Shared Cardiac Representation
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2608.12944
   Electrocardiography (ECG), photoplethysmography (PPG), and phonocardiography (PCG) provide complementary views of the same cardiac cycle, yet existing cardiac foundation models are trained for a single sensing modality, leaving the shared physiology across sensors unexploited.
6. Energy Efficiency Maximization for FAS-Assisted Downlink Communication in Mobile Embodied AI Networks (MEAN) over Interference Channels
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12849
   In this paper, we investigate a fluid antenna system (FAS)-assisted downlink mobile embodied AI network (MEAN) over interference channels, where multiple base station (BS)-agent pairs reuse the same spectrum.
7. SraVaani 1.0: Scaling Inclusive Speech Recognition for Indic Languages
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.08235
   India's linguistic landscape spans over 700 languages and thousands of dialects, yet the vast majority of automatic speech recognition (ASR) systems support only a small fraction of this diversity.
8. Resource-efficient Semantic Coding Schemes with Manifold-constrained Hyper-connections
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2608.13253
   Semantic communication (SemCom) and task-oriented communication (TOC) can reduce wireless resource consumption by focusing on transmitting semantic or task-relevant information instead of raw messages.
9. Antenna Positioning and Beamforming Optimization in MA Enabled Secure ISAC Systems: A Gradient-Based Meta Learning Approach
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12870
   Integrated sensing and communications (ISAC) significantly improves spectral efficiency but introduces security risks regarding the interception of embedded communication signals.
10. Longest Filled Common Subsequence for Song Identification from Degraded Audio via Construct--Merge--Solve--Adapt Optimization
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2509.12261
   This paper addresses the Longest Filled Common Subsequence (LFCS) problem, a challenging NP-hard problem with applications in bioinformatics, including gene mutation prediction and genomic data reconstruction.
11. Can Generalist Vision Language Models (VLMs) Rival Specialist Medical VLMs? Benchmarking and Strategic Insights
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2506.17337
   Vision Language Models (VLMs) have shown promise in automating image diagnosis and interpretation in clinical settings.
12. Efficient and Accurate Surrogate-Assisted Electromagnetic Parameter Calibration for 6G Digital Twin Channels
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12872
   High-fidelity ray-tracing (RT) channel reconstruction is a fundamental step toward building digital twins for the era of 6G wireless communications.
13. Hybrid Real- and Complex-Valued Neural Network Architecture for Speech Enhancement
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2509.21185
   This paper investigates hybrid real- and complex-valued neural networks for monaural speech enhancement.
14. Improving Deep Learning for Accelerated MRI With Data Filtering
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2508.13822
   Deep neural networks achieve state-of-the-art results for accelerated MRI reconstruction.
15. GML-Based Optimization for Movable Antenna Wireless Networks: Challenges and Opportunities
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12882
   Movable antenna (MA) is proposed as an emerging technology for future wireless networks.
16. Cloud-Boosted Low-Compute Multi-Channel Speech Enhancement
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.07423
   Low-latency, low-compute speech enhancement is essential for wearable devices with real-time communication requirements, but strict computational constraints significantly limit on-device performance.
17. Learning Class Difficulty via Dynamic Focal Attention for Histopathology Segmentation
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2604.13479
   Frequency-based loss reweighting, the standard remedy for imbalanced histopathology segmentation, implicitly assumes that rare classes are difficult.
18. Input-Correlated Supervision Noise Limits the Benefits of OTA Training for Learned Receivers
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.12918
   While learned wireless receivers are typically studied using synthetic data, the impact of over-the-air (OTA) measurements for training remains unclear.
19. Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.12082
   Language model (LM)-based speech enhancement (SE) has recently emerged rapidly using latent space features of neural audio codecs (NACs).
20. Exploring Sparsity for Parameter Efficient Fine Tuning Using Wavelets for Vision
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2505.12532
   Efficiently adapting large pretrained models is critical under tight compute and memory budgets.
21. Wideband Large-Array Processing and Sparse Design for Angle Imaging
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.13005
   This paper shows that wideband large-array processing can recover a large number of angle pixels with far fewer antenna elements.
22. On-Policy Self-Distillation for Multi-Dialect ASR: Mastering Dialects, Retaining Mandarin
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.11898
   Recent large-scale ASR models already achieve strong Mandarin recognition accuracy and have some ability to recognize Chinese dialects.
23. An Information Theory Analysis of Whole Slide Image Pathology AI and Diagnostic Field Selection AI Under Limited Resources
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2608.10846
   A key issue in using AI for pathology diagnosis is what image information should be given to the AI and how limited analysis resources should be used.
24. Analysis and Approximation of a Spatially Wideband Antenna Array Factor
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.13125
   This work investigates the spatially wideband (SWB) antenna array factor (AF) of uniform linear arrays.
25. Deep Learning Based Relative Transfer Matrix Estimation for Multiple Sources and Multiple Microphones
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.11627
   The Relative Transfer Matrix (ReTM), recently introduced as a generalization of the relative transfer function for multiple receivers and sources, shows promising performance when applied to speech enhancement in noisy environments.
26. Quantization-Aware Neuromorphic Architecture for Skin Lesion Classification on Resource-Constrained Devices
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2507.15958
   On-device skin lesion analysis is constrained by the compute and energy cost of conventional CNN inference and by the need for lightweight calibration under clinical data shift.
27. Exploiting Phase Noise for Sensing Privacy in ISAC Systems
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.13270
   We investigate sensing privacy in orthogonal frequency-division multiplexing (OFDM) integrated sensing and communication (ISAC) systems under the impact of phase noise (PN) arising from local oscillator (LO) imperfections.
28. Beyond Reconstruction: Full-Context Generative DiT for Music Generation
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2608.08787
   Hybrid music generators combine the long-range planning of an autoregressive language model with the fidelity of a diffusion- or flow-based acoustic renderer.
29. RankByGene: Gene-Guided Histopathology Representation Learning Through Cross-Modal Ranking Consistency
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2411.15076
   Spatial transcriptomics (ST) provides essential spatial context by mapping gene expression within tissue, enabling detailed study of cellular heterogeneity and tissue organization.
30. Deterministic Maximum Likelihood Direction Finding in the Mixture Noise of Gaussian and Spherically Invariant Components
   Source: eess.SP (Signal Processing)
   Link: https://arxiv.org/abs/2608.13294
   Spherically invariant (SI) random processes can model impulsive noise and unreliable measurements.
31. Cyclostationarity Analysis as a Complement to Self-Supervised Representations for Speech Deepfake Detection
   Source: eess.AS (Audio and Speech Processing)
   Link: https://arxiv.org/abs/2603.03921
   Speech deepfake detection (SDD) is essential for maintaining trust in voice-driven technologies and digital media.
32. Soft-Attention Improves Skin Cancer Classification Performance
   Source: eess.IV (Image and Video Processing)
   Link: https://arxiv.org/abs/2105.03358
   In clinical applications, neural networks must focus on and highlight the most important parts of an input image.

Sources in this brief: eess.AS (Audio and Speech Processing); eess.IV (Image and Video Processing); eess.SP (Signal Processing).
Selected 32 of 424 available items for this weekly brief.