Woon‐Seng Gan

Kenneth Ooi, Bhan Lam, Joo Young Hong et al. · gatech

The ecological validity of soundscape studies usually rests on a choice of soundscapes that are representative of the perceptual space under investigation. For example, a soundscape pleasantness study might investigate locations with soundscapes ranging from "pleasant" to "annoying". The choice of soundscapes is typically researcher-led, but a participant-led process can reduce selection bias and improve result reliability. Hence, we propose a robust participant-led method to pinpoint characteristic soundscapes possessing arbitrary perceptual attributes. We validate our method by identifying Singaporean soundscapes spanning the perceptual quadrants generated from the "Pleasantness" and "Eventfulness" axes of the ISO 12913-2 circumplex model of soundscape perception, as perceived by local experts. From memory and experience, 67 participants first selected locations corresponding to each perceptual quadrant in each major planning region of Singapore. We then performed weighted k-means clustering on the selected locations, with weights for each location derived from previous frequencies and durations spent in each location by each participant. Weights hence acted as proxies for participant confidence. In total, 62 locations were thereby identified as suitable locations with characteristic soundscapes for further research utilizing the ISO 12913-2 perceptual quadrants. Audio-visual recordings and acoustic characterization of the soundscapes will be made in a future study.

Karn N. Watcharasupat, Kenneth Ooi, Bhan Lam et al. · gatech

The selection of maskers and playback gain levels in a soundscape augmentation system is crucial to its effectiveness in improving the overall acoustic comfort of a given environment. Traditionally, the selection of appropriate maskers and gain levels has been informed by expert opinion, which may not representative of the target population, or by listening tests, which can be time-consuming and labour-intensive. Furthermore, the resulting static choices of masker and gain are often inflexible to the dynamic nature of real-world soundscapes. In this work, we utilized a deep learning model to perform joint selection of the optimal masker and its gain level for a given soundscape. The proposed model was designed with highly modular building blocks, allowing for an optimized inference process that can quickly search through a large number of masker and gain combinations. In addition, we introduced the use of feature-domain soundscape augmentation conditioned on the digital gain level, eliminating the computationally expensive waveform-domain mixing process during inference time, as well as the tedious pre-calibration process required for new maskers. The proposed system was validated on a large-scale dataset of subjective responses to augmented soundscapes with more than 440 participants, ensuring the ability of the model to predict combined effect of the masker and its gain level on the perceptual pleasantness level.

Zhengding Luo, Dongyuan Shi, Woon-Seng Gan

The selective fixed-filter active noise control (SFANC) method selecting the best pre-trained control filters for various types of noise can achieve a fast response time. However, it may lead to large steady-state errors due to inaccurate filter selection and the lack of adaptability. In comparison, the filtered-X normalized least-mean-square (FxNLMS) algorithm can obtain lower steady-state errors through adaptive optimization. Nonetheless, its slow convergence has a detrimental effect on dynamic noise attenuation. Therefore, this paper proposes a hybrid SFANC-FxNLMS approach to overcome the adaptive algorithm's slow convergence and provide a better noise reduction level than the SFANC method. A lightweight one-dimensional convolutional neural network (1D CNN) is designed to automatically select the most suitable pre-trained control filter for each frame of the primary noise. Meanwhile, the FxNLMS algorithm continues to update the coefficients of the chosen pre-trained control filter at the sampling rate. Owing to the effective combination of the two algorithms, experimental results show that the hybrid SFANC-FxNLMS algorithm can achieve a rapid response time, a low noise reduction error, and a high degree of robustness.

Zhengding Luo, Dongyuan Shi, Xiaoyi Shen et al.

Due to the slow convergence and poor tracking ability, conventional LMS-based adaptive algorithms are less capable of handling dynamic noises. Selective fixed-filter active noise control (SFANC) can significantly reduce response time by selecting appropriate pre-trained control filters for different noises. Nonetheless, the limited number of pre-trained control filters may affect noise reduction performance, especially when the incoming noise differs much from the initial noises during pre-training. Therefore, a generative fixed-filter active noise control (GFANC) method is proposed in this paper to overcome the limitation. Based on deep learning and a perfect-reconstruction filter bank, the GFANC method only requires a few prior data (one pre-trained broadband control filter) to automatically generate suitable control filters for various noises. The efficacy of the GFANC method is demonstrated by numerical simulations on real-recorded noises.

Zhengding Luo, Dongyuan Shi, Woon-Seng Gan

Due to its rapid response time and a high degree of robustness, the selective fixed-filter active noise control (SFANC) method appears to be a viable candidate for widespread use in a variety of practical active noise control (ANC) systems. In comparison to conventional fixed-filter ANC methods, SFANC can select the pre-trained control filters for different types of noise. Deep learning technologies, thus, can be used in SFANC methods to enable a more flexible selection of the most appropriate control filters for attenuating various noises. Furthermore, with the assistance of a deep neural network, the selecting strategy can be learned automatically from noise data rather than through trial and error, which significantly simplifies and improves the practicability of ANC design. Therefore, this paper investigates the performance of SFANC based on different one-dimensional and two-dimensional convolutional neural networks. Additionally, we conducted comparative analyses of several network training strategies and discovered that fine-tuning could improve selection performance.

Jin Jie Sean Yeo, Ee-Leng Tan, Jisheng Bai et al.

In this technical report, we describe the SNTL-NTU team's submission for Task 1 Data-Efficient Low-Complexity Acoustic Scene Classification of the detection and classification of acoustic scenes and events (DCASE) 2024 challenge. Three systems are introduced to tackle training splits of different sizes. For small training splits, we explored reducing the complexity of the provided baseline model by reducing the number of base channels. We introduce data augmentation in the form of mixup to increase the diversity of training samples. For the larger training splits, we use FocusNet to provide confusing class information to an ensemble of multiple Patchout faSt Spectrogram Transformer (PaSST) models and baseline models trained on the original sampling rate of 44.1 kHz. We use Knowledge Distillation to distill the ensemble model to the baseline student model. Training the systems on the TAU Urban Acoustic Scene 2022 Mobile development dataset yielded the highest average testing accuracy of (62.21, 59.82, 56.81, 53.03, 47.97)% on split (100, 50, 25, 10, 5)% respectively over the three systems.

Haowen Li, Zhengding Luo, Dongyuan Shi et al.

Direction-of-Arrival (DOA) estimation is critical in spatial audio and acoustic signal processing, with wide-ranging applications in real-world. Most existing DOA models are trained on synthetic data by convolving clean speech with room impulse responses (RIRs), which limits their generalizability due to constrained acoustic diversity. In this paper, we revisit DOA estimation using a recently introduced dataset constructed with the assistance of large language models (LLMs), which provides more realistic and diverse spatial audio scenes. We benchmark several representative neural-based DOA methods on this dataset and propose LightDOA, a lightweight DOA estimation model based on depthwise separable convolutions, specifically designed for mutil-channel input in varying environments. Experimental results show that LightDOA achieves satisfactory accuracy and robustness across various acoustic scenes while maintaining low computational complexity. This study not only highlights the potential of spatial audio synthesized with the assistance of LLMs in advancing robust and efficient DOA estimation research, but also highlights LightDOA as efficient solution for resource-constrained applications.

Jing Yang Lee, Kong Aik Lee, Woon-Seng Gan

To engage human users in meaningful conversation, open-domain dialogue agents are required to generate diverse and contextually coherent dialogue. Despite recent advancements, which can be attributed to the usage of pretrained language models, the generation of diverse and coherent dialogue remains an open research problem. A popular approach to address this issue involves the adaptation of variational frameworks. However, while these approaches successfully improve diversity, they tend to compromise on contextual coherence. Hence, we propose the Bayesian Open-domain Dialogue with Empirical Bayes (BODEB) framework, an empirical bayes framework for constructing an Bayesian open-domain dialogue agent by leveraging pretrained parameters to inform the prior and posterior parameter distributions. Empirical results show that BODEB achieves better results in terms of both diversity and coherence compared to variational frameworks.

Jing Yang Lee, Kong Aik Lee, Woon-Seng Gan

Despite recent progress in generative open-domain dialogue, the issue of low response diversity persists. Prior works have addressed this issue via either novel objective functions, alternative learning approaches such as variational frameworks, or architectural extensions such as the Randomized Link (RL) Transformer. However, these approaches typically entail either additional difficulties during training/inference, or a significant increase in model size and complexity. Hence, we propose the \underline{Pa}rtially \underline{Ra}ndomized trans\underline{Former} (PaRaFormer), a simple extension of the transformer which involves freezing the weights of selected layers after random initialization. Experimental results reveal that the performance of the PaRaformer is comparable to that of the aforementioned approaches, despite not entailing any additional training difficulty or increase in model complexity.

Jisheng Bai, Mou Wang, Haohe Liu et al.

Acoustic scene classification (ASC) is a crucial research problem in computational auditory scene analysis, and it aims to recognize the unique acoustic characteristics of an environment. One of the challenges of the ASC task is the domain shift between training and testing data. Since 2018, ASC challenges have focused on the generalization of ASC models across different recording devices. Although this task, in recent years, has achieved substantial progress in device generalization, the challenge of domain shift between different geographical regions, involving discrepancies such as time, space, culture, and language, remains insufficiently explored at present. In addition, considering the abundance of unlabeled acoustic scene data in the real world, it is important to study the possible ways to utilize these unlabelled data. Therefore, we introduce the task Semi-supervised Acoustic Scene Classification under Domain Shift in the ICME 2024 Grand Challenge. We encourage participants to innovate with semi-supervised learning techniques, aiming to develop more robust ASC models under domain shift.

Zhengding Luo, Dongyuan Shi, Xiaoyi Shen et al.

Delayless noise control is achieved by our earlier generative fixed-filter active noise control (GFANC) framework through efficient coordination between the co-processor and real-time controller. However, the one-dimensional convolutional neural network (1D CNN) in the co-processor requires initial training using labelled noise datasets. Labelling noise data can be resource-intensive and may introduce some biases. In this paper, we propose an unsupervised-GFANC approach to simplify the 1D CNN training process and enhance its practicality. During training, the co-processor and real-time controller are integrated into an end-to-end differentiable ANC system. This enables us to use the accumulated squared error signal as the loss for training the 1D CNN. With this unsupervised learning paradigm, the unsupervised-GFANC method not only omits the labelling process but also exhibits better noise reduction performance compared to the supervised GFANC method in real noise experiments.

Jing Yang Lee, Kong-Aik Lee, Woon-Seng Gan

Open-domain Dialogue (OD) exhibits a one-to-many (o2m) property, whereby multiple appropriate responses exist for a single dialogue context. Despite prior research showing that modeling this property boosts response diversity, most modern LLM-based dialogue agents do not explicitly do so. In this work, we model the o2m property of OD in LLMs by decomposing OD generation into two key tasks: Multi-Response Generation (MRG) and Preference-based Selection (PS), which entail generating a set of n semantically and lexically diverse high-quality responses for a given dialogue context, followed by selecting a single response based on human preference, respectively. To facilitate MRG and PS, we introduce o2mDial, a dialogue corpus explicitly designed to capture the o2m property by featuring multiple plausible responses for each context. Leveraging o2mDial, we propose new in-context learning and instruction-tuning strategies, as well as novel evaluation metrics for MRG, alongside a model-based approach for PS. Empirical results demonstrate that applying the proposed two-stage framework to smaller LLMs for OD generation enhances overall response diversity while maintaining contextual coherence, improving response quality by up to 90%, bringing them closer to the performance of larger models.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

In recent years, latent variable models, such as the Conditional Variational Auto Encoder (CVAE), have been applied to both personalized and empathetic dialogue generation. Prior work have largely focused on generating diverse dialogue responses that exhibit persona consistency and empathy. However, when it comes to the contextual coherence of the generated responses, there is still room for improvement. Hence, to improve the contextual coherence, we propose a novel Uncertainty Aware CVAE (UA-CVAE) framework. The UA-CVAE framework involves approximating and incorporating the aleatoric uncertainty during response generation. We apply our framework to both personalized and empathetic dialogue generation. Empirical results show that our framework significantly improves the contextual coherence of the generated response. Additionally, we introduce a novel automatic metric for measuring contextual coherence, which was found to correlate positively with human judgement.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

The generation of personalized dialogue is vital to natural and human-like conversation. Typically, personalized dialogue generation models involve conditioning the generated response on the dialogue history and a representation of the persona/personality of the interlocutor. As it is impractical to obtain the persona/personality representations for every interlocutor, recent works have explored the possibility of generating personalized dialogue by finetuning the model with dialogue examples corresponding to a given persona instead. However, in real-world implementations, a sufficient number of corresponding dialogue examples are also rarely available. Hence, in this paper, we propose a Dual Latent Variable Generator (DLVGen) capable of generating personalized dialogue in the absence of any persona/personality information or any corresponding dialogue examples. Unlike prior work, DLVGen models the latent distribution over potential responses as well as the latent distribution over the agent's potential persona. During inference, latent variables are sampled from both distributions and fed into the decoder. Empirical results show that DLVGen is capable of generating diverse responses which accurately incorporate the agent's persona.

Thi Ngoc Tho Nguyen, Douglas L. Jones, Karn N. Watcharasupat et al.

Polyphonic sound event localization and detection (SELD) has many practical applications in acoustic sensing and monitoring. However, the development of real-time SELD has been limited by the demanding computational requirement of most recent SELD systems. In this work, we introduce SALSA-Lite, a fast and effective feature for polyphonic SELD using microphone array inputs. SALSA-Lite is a lightweight variation of a previously proposed SALSA feature for polyphonic SELD. SALSA, which stands for Spatial Cue-Augmented Log-Spectrogram, consists of multichannel log-spectrograms stacked channelwise with the normalized principal eigenvectors of the spectrotemporally corresponding spatial covariance matrices. In contrast to SALSA, which uses eigenvector-based spatial features, SALSA-Lite uses normalized inter-channel phase differences as spatial features, allowing a 30-fold speedup compared to the original SALSA feature. Experimental results on the TAU-NIGENS Spatial Sound Events 2021 dataset showed that the SALSA-Lite feature achieved competitive performance compared to the full SALSA feature, and significantly outperformed the traditional feature set of multichannel log-mel spectrograms with generalized cross-correlation spectra. Specifically, using SALSA-Lite features increased localization-dependent F1 score and class-dependent localization recall by 15% and 5%, respectively, compared to using multichannel log-mel spectrograms with generalized cross-correlation spectra.

Kenneth Ooi, Karn N. Watcharasupat, Santi Peksi et al.

This paper introduces SINGA:PURA, a strongly labelled polyphonic urban sound dataset with spatiotemporal context. The data were collected via several recording units deployed across Singapore as a part of a wireless acoustic sensor network. These recordings were made as part of a project to identify and mitigate noise sources in Singapore, but also possess a wider applicability to sound event detection, classification, and localization. This paper introduces an accompanying hierarchical label taxonomy, which has been designed to be compatible with other existing datasets for urban sound tagging while also able to capture sound events unique to the Singaporean context. This paper details the data collection, annotation, and processing methodologies for the creation of the dataset. We further perform exploratory data analysis and include the performance of a baseline model on the dataset as a benchmark.

Shulin Wen, Duy Hai Nguyen, Miqing Wang et al.

Construction workers and residents live near around construction sites are exposed to noises that might cause hearing loss, high blood pressure, heart disease, sleep disturbance and stress. Regulations has been carried out by national governments to limit the maximum permissible noise levels for construction works. A four-channel active noise control system mounted on the opening of an enclosure is designed to prevent the machinery noise from spreading around and retaining the heat diffusion path. Multi-channel FxLMS algorithm in time domain is implemented on the main controller. A Genelec speaker is placed inside the box as the primary noise source to play back different types of noises. Analyses and experiments are carried out to investigate the controllable frequency range of this ANC system in detail. Considerable noise reduction performance is achieved for different recorded practical construction noises.

Karn N. Watcharasupat, Thi Ngoc Tho Nguyen, Woon-Seng Gan et al.

Echo and noise suppression is an integral part of a full-duplex communication system. Many recent acoustic echo cancellation (AEC) systems rely on a separate adaptive filtering module for linear echo suppression and a neural module for residual echo suppression. However, not only do adaptive filtering modules require convergence and remain susceptible to changes in acoustic environments, but this two-stage framework also often introduces unnecessary delays to the AEC system when neural modules are already capable of both linear and nonlinear echo suppression. In this paper, we exploit the offset-compensating ability of complex time-frequency masks and propose an end-to-end complex-valued neural network architecture. The building block of the proposed model is a pseudocomplex extension based on the densely-connected multidilated DenseNet (D3Net) building block, resulting in a very small network of only 354K parameters. The architecture utilized the multi-resolution nature of the D3Net building blocks to eliminate the need for pooling, allowing the network to extract features using large receptive fields without any loss of output resolution. We also propose a dual-mask technique for joint echo and noise suppression with simultaneous speech enhancement. Evaluation on both synthetic and real test sets demonstrated promising results across multiple energy-based metrics and perceptual proxies.

Thi Ngoc Tho Nguyen, Karn N. Watcharasupat, Ngoc Khanh Nguyen et al.

Sound event localization and detection (SELD) consists of two subtasks, which are sound event detection and direction-of-arrival estimation. While sound event detection mainly relies on time-frequency patterns to distinguish different sound classes, direction-of-arrival estimation uses amplitude and/or phase differences between microphones to estimate source directions. As a result, it is often difficult to jointly optimize these two subtasks. We propose a novel feature called Spatial cue-Augmented Log-SpectrogrAm (SALSA) with exact time-frequency mapping between the signal power and the source directional cues, which is crucial for resolving overlapping sound sources. The SALSA feature consists of multichannel log-spectrograms stacked along with the normalized principal eigenvector of the spatial covariance matrix at each corresponding time-frequency bin. Depending on the microphone array format, the principal eigenvector can be normalized differently to extract amplitude and/or phase differences between the microphones. As a result, SALSA features are applicable for different microphone array formats such as first-order ambisonics (FOA) and multichannel microphone array (MIC). Experimental results on the TAU-NIGENS Spatial Sound Events 2021 dataset with directional interferences showed that SALSA features outperformed other state-of-the-art features. Specifically, the use of SALSA features in the FOA format increased the F1 score and localization recall by 6% each, compared to the multichannel log-mel spectrograms with intensity vectors. For the MIC format, using SALSA features increased F1 score and localization recall by 16% and 7%, respectively, compared to using multichannel log-mel spectrograms with generalized cross-correlation spectra.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

Conventional approaches to personalized dialogue generation typically require a large corpus, as well as predefined persona information. However, in a real-world setting, neither a large corpus of training data nor persona information are readily available. To address these practical limitations, we propose a novel multi-task meta-learning approach which involves training a model to adapt to new personas without relying on a large corpus, or on any predefined persona information. Instead, the model is tasked with generating personalized responses based on only the dialogue context. Unlike prior work, our approach leverages on the provided persona information only during training via the introduction of an auxiliary persona reconstruction task. In this paper, we introduce 2 frameworks that adopt the proposed multi-task meta-learning approach: the Multi-Task Meta-Learning (MTML) framework, and the Alternating Multi-Task Meta-Learning (AMTML) framework. Experimental results show that utilizing MTML and AMTML results in dialogue responses with greater persona consistency.

Karn N. Watcharasupat, Thi Ngoc Tho Nguyen, Ngoc Khanh Nguyen et al.

The Sørensen--Dice Coefficient has recently seen rising popularity as a loss function (also known as Dice loss) due to its robustness in tasks where the number of negative samples significantly exceeds that of positive samples, such as semantic segmentation, natural language processing, and sound event detection. Conventional training of polyphonic sound event detection systems with binary cross-entropy loss often results in suboptimal detection performance as the training is often overwhelmed by updates from negative samples. In this paper, we investigated the effect of the Dice loss, intra- and inter-modal transfer learning, data augmentation, and recording formats, on the performance of polyphonic sound event detection systems with multichannel inputs. Our analysis showed that polyphonic sound event detection systems trained with Dice loss consistently outperformed those trained with cross-entropy loss across different training settings and recording formats in terms of F1 score and error rate. We achieved further performance gains via the use of transfer learning and an appropriate combination of different data augmentation techniques.

Thi Ngoc Tho Nguyen, Karn N. Watcharasupat, Zhen Jian Lee et al.

Sound event localization and detection (SELD) is an emerging research topic that aims to unify the tasks of sound event detection and direction-of-arrival estimation. As a result, SELD inherits the challenges of both tasks, such as noise, reverberation, interference, polyphony, and non-stationarity of sound sources. Furthermore, SELD often faces an additional challenge of assigning correct correspondences between the detected sound classes and directions of arrival to multiple overlapping sound events. Previous studies have shown that unknown interferences in reverberant environments often cause major degradation in the performance of SELD systems. To further understand the challenges of the SELD task, we performed a detailed error analysis on two of our SELD systems, which both ranked second in the team category of DCASE SELD Challenge, one in 2020 and one in 2021. Experimental results indicate polyphony as the main challenge in SELD, due to the difficulty in detecting all sound events of interest. In addition, the SELD systems tend to make fewer errors for the polyphonic scenario that is dominant in the training set.

Thi Ngoc Tho Nguyen, Karn Watcharasupat, Ngoc Khanh Nguyen et al.

Sound event localization and detection consists of two subtasks which are sound event detection and direction-of-arrival estimation. While sound event detection mainly relies on time-frequency patterns to distinguish different sound classes, direction-of-arrival estimation uses magnitude or phase differences between microphones to estimate source directions. Therefore, it is often difficult to jointly train these two subtasks simultaneously. We propose a novel feature called spatial cue-augmented log-spectrogram (SALSA) with exact time-frequency mapping between the signal power and the source direction-of-arrival. The feature includes multichannel log-spectrograms stacked along with the estimated direct-to-reverberant ratio and a normalized version of the principal eigenvector of the spatial covariance matrix at each time-frequency bin on the spectrograms. Experimental results on the DCASE 2021 dataset for sound event localization and detection with directional interference showed that the deep learning-based models trained on this new feature outperformed the DCASE challenge baseline by a large margin. We combined several models with slightly different architectures that were trained on the new feature to further improve the system performances for the DCASE sound event localization and detection challenge.

T. N. T. Nguyen, D. L. Jones, R. Ranjan et al.

Sound event detection and sound event localization requires different features from audio input signals. While sound event detection mainly relies on time-frequency patterns to distinguish different event classes, sound event localization uses magnitude or phase differences between microphones to estimate source directions. Therefore, we propose a two-step system to do sound event localization and detection. In the first step, we detect the sound events and estimate the directions-of-arrival separately. In the second step, we combine the results of the event detector and direction-of-arrival estimator together. The obtained results show a significant improvement over the baseline solution for sound event localization and detection in DCASE 2019 task 3 challenge. Using the evaluation dataset, the proposed system achieved an F1 score of 93.4% for sound event detection and an error of 5.4 degrees for direction-of-arrival estimation, while the winning solution achieved an F1 score of 94.7% and an angle error of 3.7 degrees respectively.