Recognizing Ornaments in Vocal Indian Art Music with Active Annotation

Ornamentations, embellishments, or microtonal inflections are essential to melodic expression across many musical traditions, adding depth, nuance, and emotional impact to performances. Recognizing ornamentations in singing voices is key to MIR, with potential applications in music pedagogy, singer identification, genre classification, and controlled singing voice generation. However, the lack of annotated datasets and specialized modeling approaches remains a major obstacle for progress in this research area. In this work, we introduce R\=aga Ornamentation Detection (ROD), a novel dataset comprising Indian classical music recordings curated by expert musicians. The dataset is annotated using a custom Human-in-the-Loop tool for six vocal ornaments marked as event-based labels. Using this dataset, we develop an ornamentation detection model based on deep time-series analysis, preserving ornament boundaries during the chunking of long audio recordings. We conduct experiments using different train-test configurations within the ROD dataset and also evaluate our approach on a separate, manually annotated dataset of Indian classical concert recordings. Our experimental results support the superior performance of our proposed approach over the baseline CRNN.

Wavelet-Filtering of Symbolic Music Representations for Folk Tune Segmentation and Classification

The aim of this study is to evaluate a machine-learning method in which symbolic representations of folk songs are segmented and classified into tune families with Haar-wavelet filtering. The method is compared with previously proposed Gestalt-based method. Melodies are represented as discrete symbolic pitch-time signals. We apply the continuous wavelet transform (CWT) with the Haar wavelet at specific scales, obtaining filtered versions of melodies emphasizing their information at particular time-scales. We use the filtered signal for representation and segmentation, using the wavelet coefficients' local maxima to indicate local boundaries and classify segments by means of k-nearest neighbours based on standard vector-metrics (Euclidean, cityblock), and compare the results to a Gestalt-based segmentation method and metrics applied directly to the pitch signal. We found that the wavelet based segmentation and wavelet-filtering of the pitch signal lead to better classification accuracy in cross-validated evaluation when the time-scale and other parameters are optimized.

SMSAT: A Multimodal Acoustic Dataset and Deep Contrastive Learning Framework for Affective and Physiological Modeling of Spiritual Meditation

Understanding how auditory stimuli influence emotional and physiological states is fundamental to advancing affective computing and mental health technologies. In this paper, we present a multimodal evaluation of the affective and physiological impacts of three auditory conditions, that is, spiritual meditation (SM), music (M), and natural silence (NS), using a comprehensive suite of biometric signal measures. To facilitate this analysis, we introduce the Spiritual, Music, Silence Acoustic Time Series (SMSAT) dataset, a novel benchmark comprising acoustic time series (ATS) signals recorded under controlled exposure protocols, with careful attention to demographic diversity and experimental consistency. To model the auditory induced states, we develop a contrastive learning based SMSAT audio encoder that extracts highly discriminative embeddings from ATS data, achieving 99.99% classification accuracy in interclass and intraclass evaluations. Furthermore, we propose the Calmness Analysis Model (CAM), a deep learning framework integrating 25 handcrafted and learned features for affective state classification across auditory conditions, attaining robust 99.99% classification accuracy. In contrast, pairwise t tests reveal significant deviations in cardiac response characteristics (CRC) between SM analysis via ANOVA inducing more significant physiological fluctuations. Compared to existing state of the art methods reporting accuracies up to 90%, the proposed model demonstrates substantial performance gains (up to 99%). This work contributes a validated multimodal dataset and a scalable deep learning framework for affective computing applications in stress monitoring, mental well-being, and therapeutic audio-based interventions.

Bayesian Sensing for Time-Varying Channels in ISAC Systems

Future mobile networks are projected to support integrated sensing and communications in high-speed communication scenarios. Nevertheless, large Doppler shifts induced by time-varying channels may cause severe inter-carrier interference (ICI). Frequency domain shows the potential of reducing ISAC complexity as compared with other domains. However, parameter mismatching issue still exists for such sensing. In this paper, we develop a novel sensing scheme based on sparse Bayesian framework, where the delay and Doppler estimation problem in time-varying channels is formulated as a 3D multiple measurement-sparse signal recovery (MM-SSR) problem. We then propose a novel two-layer variational Bayesian inference (VBI) method to decompose the 3D MM-SSR problem into two layers and estimate the Doppler in the first layer and the delay in the second layer alternatively. Subsequently, as is benefited from newly unveiled signal construction, a simplified two-stage multiple signal classification (MUSIC)-based VBI method is proposed, where the delay and the Doppler are estimated by MUSIC and VBI, respectively. Additionally, the Cram\'er-Rao bound (CRB) of the considered sensing parameters is derived to characterize the lower bound for the proposed estimators. Corroborated by extensive simulation results, our proposed method can achieve improved mean square error (MSE) than its conventional counterparts and is robust against the target number and target speed, thereby validating its wide applicability and advantages over prior arts.

Progressive Rock Music Classification

This study investigates the classification of progressive rock music, a genre characterized by complex compositions and diverse instrumentation, distinct from other musical styles. Addressing this Music Information Retrieval (MIR) task, we extracted comprehensive audio features, including spectrograms, Mel-Frequency Cepstral Coefficients (MFCCs), chromagrams, and beat positions from song snippets using the Librosa library. A winner-take-all voting strategy was employed to aggregate snippet-level predictions into final song classifications. We conducted a comparative analysis of various machine learning techniques. Ensemble methods, encompassing Bagging (Random Forest, ExtraTrees, Bagging Classifier) and Boosting (XGBoost, Gradient Boosting), were explored, utilizing Principal Component Analysis (PCA) for dimensionality reduction to manage computational constraints with high-dimensional feature sets. Additionally, deep learning approaches were investigated, including the development of custom 1D Convolutional Neural Network (1D CNN) architectures (named "Zuck" and "Satya") featuring specific layer configurations, normalization, and activation functions. Furthermore, we fine-tuned a state-of-the-art Audio Spectrogram Transformer (AST) model, leveraging its attention-based mechanisms for audio classification. Performance evaluation on validation and test sets revealed varying effectiveness across models, with ensemble methods like Extra Trees achieving test accuracies up to 76.38%. This research provides insights into the application and relative performance of diverse machine learning paradigms for the nuanced task of progressive rock genre classification.

Parametric Near-Field MMSE Channel Estimation for sub-THz XL-MIMO Systems

Accurate channel estimation is essential for reliable communication in sub-THz extremely large (XL) MIMO systems. Deploying XL-MIMO in high-frequency bands not only increases the number of antennas, but also fundamentally alters channel propagation characteristics, placing the user equipments (UE) in the radiative near-field of the base station. This paper proposes a parametric estimation method using the multiple signal classification (MUSIC) algorithm to extract UE location data from uplink pilot signals. These parameters are used to reconstruct the spatial correlation matrix, followed by an approximation of the minimum mean square error (MMSE) channel estimator. Numerical results show that the proposed method outperforms the least-squares (LS) estimator in terms of the normalized mean-square error (NMSE), even without prior UE location knowledge.

Sparse Hybrid Linear-Morphological Networks

We investigate hybrid linear-morphological networks. Recent studies highlight the inherent affinity of morphological layers to pruning, but also their difficulty in training. We propose a hybrid network structure, wherein morphological layers are inserted between the linear layers of the network, in place of activation functions. We experiment with the following morphological layers: 1) maxout pooling layers (as a special case of a morphological layer), 2) fully connected dense morphological layers, and 3) a novel, sparsely initialized variant of (2). We conduct experiments on the Magna-Tag-A-Tune (music auto-tagging) and CIFAR-10 (image classification) datasets, replacing the linear classification heads of state-of-the-art convolutional network architectures with our proposed network structure for the various morphological layers. We demonstrate that these networks induce sparsity to their linear layers, making them more prunable under L1 unstructured pruning. We also show that on MTAT our proposed sparsely initialized layer achieves slightly better performance than ReLU, maxout, and densely initialized max-plus layers, and exhibits faster initial convergence.

DSSR-Net for Super-Resolution Radar Range Profiles

High-resolution radar range profile (RRP) is crucial for accurate target recognition and scene perception. To get a high-resolution RRP, many methods have been developed, such as multiple signal classification (MUSIC), orthogonal matching pursuit (OMP), and a few deep learning-based approaches. Although they break through the Rayleigh resolution limit determined by radar signal bandwidth, these methods either get limited super-resolution capability or work well just in high signal to noise ratio (SNR) scenarios. To overcome these limitations, in this paper, an interpretable neural network for super-resolution RRP (DSSR-Net) is proposed by integrating the advantages of both model-guided and data-driven models. Specifically, DSSR-Net is designed based on a sparse representation model with dimension scaling, and then trained on a training dataset. Through dimension scaling, DSSR-Net lifts the radar signal into high-dimensional space to extract subtle features of closely spaced objects and suppress the noise of the high-dimensional features. It improves the super-resolving power of closely spaced objects and lowers the SNR requirement of radar signals compared to existing methods. The superiority of the proposed algorithm for super-resolution RRP reconstruction is verified via experiments with both synthetic and measured data.

An Efficient GPU-based Implementation for Noise Robust Sound Source Localization

Robot audition, encompassing Sound Source Localization (SSL), Sound Source Separation (SSS), and Automatic Speech Recognition (ASR), enables robots and smart devices to acquire auditory capabilities similar to human hearing. Despite their wide applicability, processing multi-channel audio signals from microphone arrays in SSL involves computationally intensive matrix operations, which can hinder efficient deployment on Central Processing Units (CPUs), particularly in embedded systems with limited CPU resources. This paper introduces a GPU-based implementation of SSL for robot audition, utilizing the Generalized Singular Value Decomposition-based Multiple Signal Classification (GSVD-MUSIC), a noise-robust algorithm, within the HARK platform, an open-source software suite. For a 60-channel microphone array, the proposed implementation achieves significant performance improvements. On the Jetson AGX Orin, an embedded device powered by an NVIDIA GPU and ARM Cortex-A78AE v8.2 64-bit CPUs, we observe speedups of 4645.1x for GSVD calculations and 8.8x for the SSL module, while speedups of 2223.4x for GSVD calculation and 8.95x for the entire SSL module on a server configured with an NVIDIA A100 GPU and AMD EPYC 7352 CPUs, making real-time processing feasible for large-scale microphone arrays and providing ample capacity for real-time processing of potential subsequent machine learning or deep learning tasks.

Teaching LLMs Music Theory with In-Context Learning and Chain-of-Thought Prompting: Pedagogical Strategies for Machines

This study evaluates the baseline capabilities of Large Language Models (LLMs) like ChatGPT, Claude, and Gemini to learn concepts in music theory through in-context learning and chain-of-thought prompting. Using carefully designed prompts (in-context learning) and step-by-step worked examples (chain-of-thought prompting), we explore how LLMs can be taught increasingly complex material and how pedagogical strategies for human learners translate to educating machines. Performance is evaluated using questions from an official Canadian Royal Conservatory of Music (RCM) Level 6 examination, which covers a comprehensive range of topics, including interval and chord identification, key detection, cadence classification, and metrical analysis. Additionally, we evaluate the suitability of various music encoding formats for these tasks (ABC, Humdrum, MEI, MusicXML). All experiments were run both with and without contextual prompts. Results indicate that without context, ChatGPT with MEI performs the best at 52%, while with context, Claude with MEI performs the best at 75%. Future work will further refine prompts and expand to cover more advanced music theory concepts. This research contributes to the broader understanding of teaching LLMs and has applications for educators, students, and developers of AI music tools alike.