Small-Footprint Keyword Spotting (SF-KWS) has gained popularity in today's landscape of smart voice-activated devices, smartphones, and Internet of Things (IoT) applications. This surge is attributed to the advancements in Deep Learning, enabling the identification of predefined words or keywords from a continuous stream of words. To implement the SF-KWS model on edge devices with low power and limited memory in real-world scenarios, a efficient Tiny Machine Learning (TinyML) framework is essential. In this study, we explore seven distinct categories of techniques namely, Model Architecture, Learning Techniques, Model Compression, Attention Awareness Architecture, Feature Optimization, Neural Network Search, and Hybrid Approaches, which are suitable for developing an SF-KWS system. This comprehensive overview will serve as a valuable resource for those looking to understand, utilize, or contribute to the field of SF-KWS. The analysis conducted in this work enables the identification of numerous potential research directions, encompassing insights from automatic speech recognition research and those specifically pertinent to the realm of spoken SF-KWS.
Contrastive Language Audio Pretraining (CLAP) is a widely-used method to bridge the gap between audio and text domains. Current CLAP methods enable sound and music retrieval in English, ignoring multilingual spoken content. To address this, we introduce general language audio pretraining (GLAP), which expands CLAP with multilingual and multi-domain abilities. GLAP demonstrates its versatility by achieving competitive performance on standard audio-text retrieval benchmarks like Clotho and AudioCaps, while significantly surpassing existing methods in speech retrieval and classification tasks. Additionally, GLAP achieves strong results on widely used sound-event zero-shot benchmarks, while simultaneously outperforming previous methods on speech content benchmarks. Further keyword spotting evaluations across 50 languages emphasize GLAP's advanced multilingual capabilities. Finally, multilingual sound and music understanding is evaluated across four languages. Checkpoints and Source: https://github.com/xiaomi-research/dasheng-glap.
On-device learning at the edge enables low-latency, private personalization with improved long-term robustness and reduced maintenance costs. Yet, achieving scalable, low-power end-to-end on-chip learning, especially from real-world sequential data with a limited number of examples, is an open challenge. Indeed, accelerators supporting error backpropagation optimize for learning performance at the expense of inference efficiency, while simplified learning algorithms often fail to reach acceptable accuracy targets. In this work, we present Chameleon, leveraging three key contributions to solve these challenges. (i) A unified learning and inference architecture supports few-shot learning (FSL), continual learning (CL) and inference at only 0.5% area overhead to the inference logic. (ii) Long temporal dependencies are efficiently captured with temporal convolutional networks (TCNs), enabling the first demonstration of end-to-end on-chip FSL and CL on sequential data and inference on 16-kHz raw audio. (iii) A dual-mode, matrix-multiplication-free compute array allows either matching the power consumption of state-of-the-art inference-only keyword spotting (KWS) accelerators or enabling $4.3\times$ higher peak GOPS. Fabricated in 40-nm CMOS, Chameleon sets new accuracy records on Omniglot for end-to-end on-chip FSL (96.8%, 5-way 1-shot, 98.8%, 5-way 5-shot) and CL (82.2% final accuracy for learning 250 classes with 10 shots), while maintaining an inference accuracy of 93.3% on the 12-class Google Speech Commands dataset at an extreme-edge power budget of 3.1 $\mu$W.
Keyword spotting (KWS) is essential for voice-driven applications, demanding both accuracy and efficiency. Traditional ASR-based KWS methods, such as greedy and beam search, explore the entire search space without explicitly prioritizing keyword detection, often leading to suboptimal performance. In this paper, we propose an effective keyword-specific KWS framework by introducing a streaming-oriented CTC-Transducer-combined frame-asynchronous system with multi-head frame-asynchronous decoding (MFA-KWS). Specifically, MFA-KWS employs keyword-specific phone-synchronous decoding for CTC and replaces conventional RNN-T with Token-and-Duration Transducer to enhance both performance and efficiency. Furthermore, we explore various score fusion strategies, including single-frame-based and consistency-based methods. Extensive experiments demonstrate the superior performance of MFA-KWS, which achieves state-of-the-art results on both fixed keyword and arbitrary keywords datasets, such as Snips, MobvoiHotwords, and LibriKWS-20, while exhibiting strong robustness in noisy environments. Among fusion strategies, the consistency-based CDC-Last method delivers the best performance. Additionally, MFA-KWS achieves a 47% to 63% speed-up over the frame-synchronous baselines across various datasets. Extensive experimental results confirm that MFA-KWS is an effective and efficient KWS framework, making it well-suited for on-device deployment.
Hate speech and abusive language are global phenomena that need socio-cultural background knowledge to be understood, identified, and moderated. However, in many regions of the Global South, there have been several documented occurrences of (1) absence of moderation and (2) censorship due to the reliance on keyword spotting out of context. Further, high-profile individuals have frequently been at the center of the moderation process, while large and targeted hate speech campaigns against minorities have been overlooked. These limitations are mainly due to the lack of high-quality data in the local languages and the failure to include local communities in the collection, annotation, and moderation processes. To address this issue, we present AfriHate: a multilingual collection of hate speech and abusive language datasets in 15 African languages. Each instance in AfriHate is annotated by native speakers familiar with the local culture. We report the challenges related to the construction of the datasets and present various classification baseline results with and without using LLMs. The datasets, individual annotations, and hate speech and offensive language lexicons are available on https://github.com/AfriHate/AfriHate
The increasing success of audio foundation models across various tasks has led to a growing need for improved interpretability to understand their intricate decision-making processes better. Existing methods primarily focus on explaining these models by attributing importance to elements within the input space based on their influence on the final decision. In this paper, we introduce a novel audio explanation method that capitalises on the generative capacity of audio foundation models. Our method leverages the intrinsic representational power of the embedding space within these models by integrating established feature attribution techniques to identify significant features in this space. The method then generates listenable audio explanations by prioritising the most important features. Through rigorous benchmarking against standard datasets, including keyword spotting and speech emotion recognition, our model demonstrates its efficacy in producing audio explanations.
The Keyword Spotting (KWS) task involves continuous audio stream monitoring to detect predefined words, requiring low energy devices for continuous processing. Neuromorphic devices effectively address this energy challenge. However, the general neuromorphic KWS pipeline, from microphone to Spiking Neural Network (SNN), entails multiple processing stages. Leveraging the popularity of Pulse Density Modulation (PDM) microphones in modern devices and their similarity to spiking neurons, we propose a direct microphone-to-SNN connection. This approach eliminates intermediate stages, notably reducing computational costs. The system achieved an accuracy of 91.54\% on the Google Speech Command (GSC) dataset, surpassing the state-of-the-art for the Spiking Speech Command (SSC) dataset which is a bio-inspired encoded GSC. Furthermore, the observed sparsity in network activity and connectivity indicates potential for remarkably low energy consumption in a neuromorphic device implementation.
We present a Spiking Neural Network (SNN) model that incorporates learnable synaptic delays through two approaches: per-synapse delay learning via Dilated Convolutions with Learnable Spacings (DCLS) and a dynamic pruning strategy that also serves as a form of delay learning. In the latter approach, the network dynamically selects and prunes connections, optimizing the delays in sparse connectivity settings. We evaluate both approaches on the Raw Heidelberg Digits keyword spotting benchmark using Backpropagation Through Time with surrogate gradients. Our analysis of the spatio-temporal structure of synaptic interactions reveals that, after training, excitation and inhibition group together in space and time. Notably, the dynamic pruning approach, which employs DEEP R for connection removal and RigL for reconnection, not only preserves these spatio-temporal patterns but outperforms per-synapse delay learning in sparse networks. Our results demonstrate the potential of combining delay learning with dynamic pruning to develop efficient SNN models for temporal data processing. Moreover, the preservation of spatio-temporal dynamics throughout pruning and rewiring highlights the robustness of these features, providing a solid foundation for future neuromorphic computing applications.
For noisy environments, ensuring the robustness of keyword spotting (KWS) systems is essential. While much research has focused on noisy KWS, less attention has been paid to multi-talker mixed speech scenarios. Unlike the usual cocktail party problem where multi-talker speech is separated using speaker clues, the key challenge here is to extract the target speech for KWS based on text clues. To address it, this paper proposes a novel Text-aware Permutation Determinization Training method for multi-talker KWS with a clue-based Speech Separation front-end (TPDT-SS). Our research highlights the critical role of SS front-ends and shows that incorporating keyword-specific clues into these models can greatly enhance the effectiveness. TPDT-SS shows remarkable success in addressing permutation problems in mixed keyword speech, thereby greatly boosting the performance of the backend. Additionally, fine-tuning our system on unseen mixed speech results in further performance improvement.
Keyword Spotting (KWS) is essential in edge computing requiring rapid and energy-efficient responses. Spiking Neural Networks (SNNs) are well-suited for KWS for their efficiency and temporal capacity for speech. To further reduce the latency and energy consumption, this study introduces ED-sKWS, an SNN-based KWS model with an early-decision mechanism that can stop speech processing and output the result before the end of speech utterance. Furthermore, we introduce a Cumulative Temporal (CT) loss that can enhance prediction accuracy at both the intermediate and final timesteps. To evaluate early-decision performance, we present the SC-100 dataset including 100 speech commands with beginning and end timestamp annotation. Experiments on the Google Speech Commands v2 and our SC-100 datasets show that ED-sKWS maintains competitive accuracy with 61% timesteps and 52% energy consumption compared to SNN models without early-decision mechanism, ensuring rapid response and energy efficiency.
With the increasing prevalence of voice-activated devices and applications, keyword spotting (KWS) models enable users to interact with technology hands-free, enhancing convenience and accessibility in various contexts. Deploying KWS models on edge devices, such as smartphones and embedded systems, offers significant benefits for real-time applications, privacy, and bandwidth efficiency. However, these devices often possess limited computational power and memory. This necessitates optimizing neural network models for efficiency without significantly compromising their accuracy. To address these challenges, we propose a novel keyword-spotting model based on sparse input representation followed by a linear classifier. The model is four times faster than the previous state-of-the-art edge device-compatible model with better accuracy. We show that our method is also more robust in noisy environments while being fast. Our code is available at: https://github.com/jsvir/sparknet.
In this paper, we propose MM-KWS, a novel approach to user-defined keyword spotting leveraging multi-modal enrollments of text and speech templates. Unlike previous methods that focus solely on either text or speech features, MM-KWS extracts phoneme, text, and speech embeddings from both modalities. These embeddings are then compared with the query speech embedding to detect the target keywords. To ensure the applicability of MM-KWS across diverse languages, we utilize a feature extractor incorporating several multilingual pre-trained models. Subsequently, we validate its effectiveness on Mandarin and English tasks. In addition, we have integrated advanced data augmentation tools for hard case mining to enhance MM-KWS in distinguishing confusable words. Experimental results on the LibriPhrase and WenetPhrase datasets demonstrate that MM-KWS outperforms prior methods significantly.
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.
There is increasing interest in the use of the LEArnable Front-end (LEAF) in a variety of speech processing systems. However, there is a dearth of analyses of what is actually learnt and the relative importance of training the different components of the front-end. In this paper, we investigate this question on keyword spotting, speech-based emotion recognition and language identification tasks and find that the filters for spectral decomposition and the low pass filter used to estimate spectral energy variations exhibit no learning and the per-channel energy normalisation (PCEN) is the key component that is learnt. Following this, we explore the potential of adapting only the PCEN layer with a small amount of noisy data to enable it to learn appropriate dynamic range compression that better suits the noise conditions. This in turn enables a system trained on clean speech to work more accurately on noisy test data as demonstrated by the experimental results reported in this paper.
Voice assistants are now widely available, and to activate them a keyword spotting (KWS) algorithm is used. Modern KWS systems are mainly trained using supervised learning methods and require a large amount of labelled data to achieve a good performance. Leveraging unlabelled data through self-supervised learning (SSL) has been shown to increase the accuracy in clean conditions. This paper explores how SSL pretraining such as Data2Vec can be used to enhance the robustness of KWS models in noisy conditions, which is under-explored. Models of three different sizes are pretrained using different pretraining approaches and then fine-tuned for KWS. These models are then tested and compared to models trained using two baseline supervised learning methods, one being standard training using clean data and the other one being multi-style training (MTR). The results show that pretraining and fine-tuning on clean data is superior to supervised learning on clean data across all testing conditions, and superior to supervised MTR for testing conditions of SNR above 5 dB. This indicates that pretraining alone can increase the model's robustness. Finally, it is found that using noisy data for pretraining models, especially with the Data2Vec-denoising approach, significantly enhances the robustness of KWS models in noisy conditions.
Keyword spotting systems for always-on TinyML-constrained applications require on-site tuning to boost the accuracy of offline trained classifiers when deployed in unseen inference conditions. Adapting to the speech peculiarities of target users requires many in-domain samples, often unavailable in real-world scenarios. Furthermore, current on-device learning techniques rely on computationally intensive and memory-hungry backbone update schemes, unfit for always-on, battery-powered devices. In this work, we propose a novel on-device learning architecture, composed of a pretrained backbone and a user-aware embedding learning the user's speech characteristics. The so-generated features are fused and used to classify the input utterance. For domain shifts generated by unseen speakers, we measure error rate reductions of up to 19% from 30.1% to 24.3% based on the 35-class problem of the Google Speech Commands dataset, through the inexpensive update of the user projections. We moreover demonstrate the few-shot learning capabilities of our proposed architecture in sample- and class-scarce learning conditions. With 23.7 kparameters and 1 MFLOP per epoch required for on-device training, our system is feasible for TinyML applications aimed at battery-powered microcontrollers.
In this project, we demonstrate that phoneme-based models for speech processing can achieve strong crosslinguistic generalizability to unseen languages. We curated the IPAPACK, a massively multilingual speech corpora with phonemic transcriptions, encompassing more than 115 languages from diverse language families, selectively checked by linguists. Based on the IPAPACK, we propose CLAP-IPA, a multi-lingual phoneme-speech contrastive embedding model capable of open-vocabulary matching between arbitrary speech signals and phonemic sequences. The proposed model was tested on 95 unseen languages, showing strong generalizability across languages. Temporal alignments between phonemes and speech signals also emerged from contrastive training, enabling zeroshot forced alignment in unseen languages. We further introduced a neural forced aligner IPA-ALIGNER by finetuning CLAP-IPA with the Forward-Sum loss to learn better phone-to-audio alignment. Evaluation results suggest that IPA-ALIGNER can generalize to unseen languages without adaptation.
Deep learning models have become widely adopted in various domains, but their performance heavily relies on a vast amount of data. Datasets often contain a large number of irrelevant or redundant samples, which can lead to computational inefficiencies during the training. In this work, we introduce, for the first time in the context of the audio domain, the k-means clustering as a method for efficient data pruning. K-means clustering provides a way to group similar samples together, allowing the reduction of the size of the dataset while preserving its representative characteristics. As an example, we perform clustering analysis on the keyword spotting (KWS) dataset. We discuss how k-means clustering can significantly reduce the size of audio datasets while maintaining the classification performance across neural networks (NNs) with different architectures. We further comment on the role of scaling analysis in identifying the optimal pruning strategies for a large number of samples. Our studies serve as a proof-of-principle, demonstrating the potential of data selection with distance-based clustering algorithms for the audio domain and highlighting promising research avenues.
This study presents a novel zero-shot user-defined keyword spotting model that utilizes the audio-phoneme relationship of the keyword to improve performance. Unlike the previous approach that estimates at utterance level, we use both utterance and phoneme level information. Our proposed method comprises a two-stream speech encoder architecture, self-attention-based pattern extractor, and phoneme-level detection loss for high performance in various pronunciation environments. Based on experimental results, our proposed model outperforms the baseline model and achieves competitive performance compared with full-shot keyword spotting models. Our proposed model significantly improves the EER and AUC across all datasets, including familiar words, proper nouns, and indistinguishable pronunciations, with an average relative improvement of 67% and 80%, respectively. The implementation code of our proposed model is available at https://github.com/ncsoft/PhonMatchNet.
Recent advances in segmentation-free keyword spotting treat this problem w.r.t. an object detection paradigm and borrow from state-of-the-art detection systems to simultaneously propose a word bounding box proposal mechanism and compute a corresponding representation. Contrary to the norm of such methods that rely on complex and large DNN models, we propose a novel segmentation-free system that efficiently scans a document image to find rectangular areas that include the query information. The underlying model is simple and compact, predicting character occurrences over rectangular areas through an implicitly learned scale map, trained on word-level annotated images. The proposed document scanning is then performed using this character counting in a cost-effective manner via integral images and binary search. Finally, the retrieval similarity by character counting is refined by a pyramidal representation and a CTC-based re-scoring algorithm, fully utilizing the trained CNN model. Experimental validation on two widely-used datasets shows that our method achieves state-of-the-art results outperforming the more complex alternatives, despite the simplicity of the underlying model.