Abstract--This paper presents the design, implementation, and evolution of a comprehensive multimodal room monitoring system that integrates synchronized video and audio processing for real-time activity recognition and anomaly detection. We describe two iterations of the system: an initial lightweight implementation using YOLOv8, ByteTrack, and Audio Spectrogram Transformer (AST), and an advanced version incorporating multi-model audio ensembles, hybrid object detection, bidirectional cross-modal attention, and multi-method anomaly detection. The evolution demonstrates significant improvements in accuracy, robustness, and industrial applicability. The advanced system combines three audio models (AST, Wav2V ec2, HuBERT) for comprehensive audio understanding, dual object detectors (YOLO and DETR) for improved accuracy, and sophisticated fusion mechanisms for enhanced cross-modal learning. Experimental evaluation shows the system's effectiveness in both general monitoring scenarios and specialized industrial safety applications, achieving real-time performance on standard hardware while maintaining high accuracy. The increasing demand for intelligent monitoring systems in smart homes, security applications, healthcare facilities, and industrial environments has driven significant research into multimodal perception systems. Traditional single-modality approaches, whether relying solely on visual or audio information, often fail to capture the complete context of activities and events. Visual systems may miss audio cues that provide crucial context, while audio-only systems lack spatial understanding. This paper presents a comprehensive journey through the development of a multimodal room monitoring system, documenting both an initial lightweight implementation and its evolution into an advanced industrial-grade solution.

Auscultation is crucial for diagnosing lung diseases. The COVID-19 pandemic has revealed the limitations of traditional, in-person lung sound assessments. To overcome these issues, advancements in digital stethoscopes and artificial intelligence (AI) have led to the development of new diagnostic methods. In this context, our study aims to use smartphone microphones to record and analyze lung sounds. We faced two major challenges: the difference in audio style between electronic stethoscopes and smartphone microphones, and the variability among patients. To address these challenges, we developed a method called Patient Domain Supervised Contrastive Learning (PD-SCL). By integrating this method with the Audio Spectrogram Transformer (AST) model, we significantly improved its performance by 2.4\% compared to the original AST model. This progress demonstrates that smartphones can effectively diagnose lung sounds, addressing inconsistencies in patient data and showing potential for broad use beyond traditional clinical settings. Our research contributes to making lung disease detection more accessible in the post-COVID-19 world.

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.

Systems for synthesizer sound matching, which automatically set the parameters of a synthesizer to emulate an input sound, have the potential to make the process of synthesizer programming faster and easier for novice and experienced musicians alike, whilst also affording new means of interaction with synthesizers. Considering the enormous variety of synthesizers in the marketplace, and the complexity of many of them, general-purpose sound matching systems that function with minimal knowledge or prior assumptions about the underlying synthesis architecture are particularly desirable. With this in mind, we introduce a synthesizer sound matching model based on the Audio Spectrogram Transformer. We demonstrate the viability of this model by training on a large synthetic dataset of randomly generated samples from the popular Massive synthesizer. We show that this model can reconstruct parameters of samples generated from a set of 16 parameters, highlighting its improved fidelity relative to multi-layer perceptron and convolutional neural network baselines. We also provide audio examples demonstrating the out-of-domain model performance in emulating vocal imitations, and sounds from other synthesizers and musical instruments.

Transformers have rapidly overtaken CNN-based architectures as the new standard in audio classification. Transformer-based models, such as the Audio Spectrogram Transformers (AST), also inherit the fixed-size input paradigm from CNNs. However, this leads to performance degradation for ASTs in the inference when input lengths vary from the training. This paper introduces an approach that enables the use of variable-length audio inputs with AST models during both training and inference. By employing sequence packing, our method ElasticAST, accommodates any audio length during training, thereby offering flexibility across all lengths and resolutions at the inference. This flexibility allows ElasticAST to maintain evaluation capabilities at various lengths or resolutions and achieve similar performance to standard ASTs trained at specific lengths or resolutions. Moreover, experiments demonstrate ElasticAST's better performance when trained and evaluated on native-length audio datasets.

Audio classification models, particularly the Audio Spectrogram Transformer (AST), play a crucial role in efficient audio analysis. However, optimizing their efficiency without compromising accuracy remains a challenge. In this paper, we introduce FastAST, a framework that integrates Token Merging (ToMe) into the AST framework. FastAST enhances inference speed without requiring extensive retraining by merging similar tokens in audio spectrograms. Furthermore, during training, FastAST brings about significant speed improvements. The experiments indicate that FastAST can increase audio classification throughput with minimal impact on accuracy. To mitigate the accuracy impact, we integrate Cross-Model Knowledge Distillation (CMKD) into the FastAST framework. Integrating ToMe and CMKD into AST results in improved accuracy compared to AST while maintaining faster inference speeds. FastAST represents a step towards real-time, resource-efficient audio analysis.

Transformers have revolutionized deep learning across various tasks, including audio representation learning, due to their powerful modeling capabilities. However, they often suffer from quadratic complexity in both GPU memory usage and computational inference time, affecting their efficiency. Recently, state space models (SSMs) like Mamba have emerged as a promising alternative, offering a more efficient approach by avoiding these complexities. Given these advantages, we explore the potential of SSM-based models in audio tasks. In this paper, we introduce Self-Supervised Audio Mamba (SSAMBA), the first self-supervised, attention-free, and SSM-based model for audio representation learning. SSAMBA leverages the bidirectional Mamba to capture complex audio patterns effectively. We incorporate a self-supervised pretraining framework that optimizes both discriminative and generative objectives, enabling the model to learn robust audio representations from large-scale, unlabeled datasets. We evaluated SSAMBA on various tasks such as audio classification, keyword spotting, and speaker identification. Our results demonstrate that SSAMBA outperforms the Self-Supervised Audio Spectrogram Transformer (SSAST) in most tasks. Notably, SSAMBA is approximately 92.7% faster in batch inference speed and 95.4% more memory-efficient than SSAST for the tiny model size with an input token size of 22k. These efficiency gains, combined with superior performance, underscore the effectiveness of SSAMBA's architectural innovation, making it a compelling choice for a wide range of audio processing applications.

Mixture of Experts (MoE) architectures have recently started burgeoning due to their ability to scale model's capacity while maintaining the computational cost affordable. Furthermore, they can be applied to both Transformers and State Space Models, the current state-of-the-art models in numerous fields. While MoE has been mostly investigated for the pre-training stage, its use in parameter-efficient transfer learning settings is under-explored. To narrow this gap, this paper attempts to demystify the use of MoE for parameter-efficient fine-tuning of Audio Spectrogram Transformers to audio and speech downstream tasks. Specifically, we propose Soft Mixture of Adapters (Soft-MoA). It exploits adapters as the experts and, leveraging the recent Soft MoE method, it relies on a soft assignment between the input tokens and experts to keep the computational time limited. Extensive experiments across 4 benchmarks demonstrate that Soft-MoA outperforms the single adapter method and performs on par with the dense MoA counterpart. We finally present ablation studies on key elements of Soft-MoA, showing for example that Soft-MoA achieves better scaling with more experts, as well as ensuring that all experts contribute to the computation of the output tokens, thus dispensing with the expert imbalance issue.

Transformers have become central to recent advances in audio classification. However, training an audio spectrogram transformer, e.g. AST, from scratch can be resource and time-intensive. Furthermore, the complexity of transformers heavily depends on the input audio spectrogram size. In this work, we aim to optimize AST training by linking to the resolution in the time-axis. We introduce multi-phase training of audio spectrogram transformers by connecting the seminal idea of coarse-to-fine with transformer models. To achieve this, we propose a set of methods for temporal compression. By employing one of these methods, the transformer model learns from lower-resolution (coarse) data in the initial phases, and then is fine-tuned with high-resolution data later in a curriculum learning strategy. Experimental results demonstrate that the proposed training mechanism for AST leads to improved (or on-par) performance with faster convergence, i.e. requiring fewer computational resources and less time. This approach is also generalizable to other AST-based methods regardless of their learning paradigms.

Respiratory sound contains crucial information for the early diagnosis of fatal lung diseases. Since the COVID-19 pandemic, there has been a growing interest in contact-free medical care based on electronic stethoscopes. To this end, cutting-edge deep learning models have been developed to diagnose lung diseases; however, it is still challenging due to the scarcity of medical data. In this study, we demonstrate that the pretrained model on large-scale visual and audio datasets can be generalized to the respiratory sound classification task. In addition, we introduce a straightforward Patch-Mix augmentation, which randomly mixes patches between different samples, with Audio Spectrogram Transformer (AST). We further propose a novel and effective Patch-Mix Contrastive Learning to distinguish the mixed representations in the latent space. Our method achieves state-of-the-art performance on the ICBHI dataset, outperforming the prior leading score by an improvement of 4.08%.