2023-06-15 更新
Feature Selection using Sparse Adaptive Bottleneck Centroid-Encoder
Authors:Tomojit Ghosh, Michael Kirby
We introduce a novel nonlinear model, Sparse Adaptive Bottleneck Centroid-Encoder (SABCE), for determining the features that discriminate between two or more classes. The algorithm aims to extract discriminatory features in groups while reconstructing the class centroids in the ambient space and simultaneously use additional penalty terms in the bottleneck layer to decrease within-class scatter and increase the separation of different class centroids. The model has a sparsity-promoting layer (SPL) with a one-to-one connection to the input layer. Along with the primary objective, we minimize the $l_{2,1}$-norm of the sparse layer, which filters out unnecessary features from input data. During training, we update class centroids by taking the Hadamard product of the centroids and weights of the sparse layer, thus ignoring the irrelevant features from the target. Therefore the proposed method learns to reconstruct the critical components of class centroids rather than the whole centroids. The algorithm is applied to various real-world data sets, including high-dimensional biological, image, speech, and accelerometer sensor data. We compared our method to different state-of-the-art feature selection techniques, including supervised Concrete Autoencoders (SCAE), Feature Selection Networks (FsNet), Stochastic Gates (STG), and LassoNet. We empirically showed that SABCE features often produced better classification accuracy than other methods on the sequester test sets, setting new state-of-the-art results.
PDF A novel nonlinear feature selection technique with new state of the art result. 22 pages (including references), 13 figures. The article is in review
点此查看论文截图
Latent Phrase Matching for Dysarthric Speech
Authors:Colin Lea, Dianna Yee, Jaya Narain, Zifang Huang, Lauren Tooley, Jeffrey P. Bigham, Leah Findlater
Many consumer speech recognition systems are not tuned for people with speech disabilities, resulting in poor recognition and user experience, especially for severe speech differences. Recent studies have emphasized interest in personalized speech models from people with atypical speech patterns. We propose a query-by-example-based personalized phrase recognition system that is trained using small amounts of speech, is language agnostic, does not assume a traditional pronunciation lexicon, and generalizes well across speech difference severities. On an internal dataset collected from 32 people with dysarthria, this approach works regardless of severity and shows a 60% improvement in recall relative to a commercial speech recognition system. On the public EasyCall dataset of dysarthric speech, our approach improves accuracy by 30.5%. Performance degrades as the number of phrases increases, but consistently outperforms ASR systems when trained with 50 unique phrases.
PDF
点此查看论文截图
I run as fast as a rabbit, can you? A Multilingual Simile Dialogue Dataset
Authors:Longxuan Ma, Weinan Zhang, Shuhan Zhou, Churui Sun, Changxin Ke, Ting Liu
A simile is a figure of speech that compares two different things (called the tenor and the vehicle) via shared properties. The tenor and the vehicle are usually connected with comparator words such as “like” or “as”. The simile phenomena are unique and complex in a real-life dialogue scene where the tenor and the vehicle can be verbal phrases or sentences, mentioned by different speakers, exist in different sentences, or occur in reversed order. However, the current simile research usually focuses on similes in a triplet tuple (tenor, property, vehicle) or a single sentence where the tenor and vehicle are usually entities or noun phrases, which could not reflect complex simile phenomena in real scenarios. In this paper, we propose a novel and high-quality multilingual simile dialogue (MSD) dataset to facilitate the study of complex simile phenomena. The MSD is the largest manually annotated simile data ($\sim$20K) and it contains both English and Chinese data. Meanwhile, the MSD data can also be used on dialogue tasks to test the ability of dialogue systems when using similes. We design 3 simile tasks (recognition, interpretation, and generation) and 2 dialogue tasks (retrieval and generation) with MSD. For each task, we provide experimental results from strong pre-trained or state-of-the-art models. The experiments demonstrate the challenge of MSD and we have released the data/code on GitHub.
PDF 13 Pages, 1 Figure, 12 Tables, ACL 2023 findings
点此查看论文截图
Learning Emotional Representations from Imbalanced Speech Data for Speech Emotion Recognition and Emotional Text-to-Speech
Authors:Shijun Wang, Jón Guðnason, Damian Borth
Effective speech emotional representations play a key role in Speech Emotion Recognition (SER) and Emotional Text-To-Speech (TTS) tasks. However, emotional speech samples are more difficult and expensive to acquire compared with Neutral style speech, which causes one issue that most related works unfortunately neglect: imbalanced datasets. Models might overfit to the majority Neutral class and fail to produce robust and effective emotional representations. In this paper, we propose an Emotion Extractor to address this issue. We use augmentation approaches to train the model and enable it to extract effective and generalizable emotional representations from imbalanced datasets. Our empirical results show that (1) for the SER task, the proposed Emotion Extractor surpasses the state-of-the-art baseline on three imbalanced datasets; (2) the produced representations from our Emotion Extractor benefit the TTS model, and enable it to synthesize more expressive speech.
PDF Accepted by INTERSPEECH2023
点此查看论文截图
Efficient Encoder-Decoder and Dual-Path Conformer for Comprehensive Feature Learning in Speech Enhancement
Authors:Junyu Wang
Current speech enhancement (SE) research has largely neglected channel attention and spatial attention, and encoder-decoder architecture-based networks have not adequately considered how to provide efficient inputs to the intermediate enhancement layer. To address these issues, this paper proposes a time-frequency (T-F) domain SE network (DPCFCS-Net) that incorporates improved densely connected blocks, dual-path modules, convolution-augmented transformers (conformers), channel attention, and spatial attention. Compared with previous models, our proposed model has a more efficient encoder-decoder and can learn comprehensive features. Experimental results on the VCTK+DEMAND dataset demonstrate that our method outperforms existing techniques in SE performance. Furthermore, the improved densely connected block and two dimensions attention module developed in this work are highly adaptable and easily integrated into existing networks.
PDF Accepted at Interspeech2023
点此查看论文截图
Record Deduplication for Entity Distribution Modeling in ASR Transcripts
Authors:Tianyu Huang, Chung Hoon Hong, Carl Wivagg, Kanna Shimizu
Voice digital assistants must keep up with trending search queries. We rely on a speech recognition model using contextual biasing with a rapidly updated set of entities, instead of frequent model retraining, to keep up with trends. There are several challenges with this approach: (1) the entity set must be frequently reconstructed, (2) the entity set is of limited size due to latency and accuracy trade-offs, and (3) finding the true entity distribution for biasing is complicated by ASR misrecognition. We address these challenges and define an entity set by modeling customers true requested entity distribution from ASR output in production using record deduplication, a technique from the field of entity resolution. Record deduplication resolves or deduplicates coreferences, including misrecognitions, of the same latent entity. Our method successfully retrieves 95% of misrecognized entities and when used for contextual biasing shows an estimated 5% relative word error rate reduction.
PDF
点此查看论文截图
Adversarial Training For Low-Resource Disfluency Correction
Authors:Vineet Bhat, Preethi Jyothi, Pushpak Bhattacharyya
Disfluencies commonly occur in conversational speech. Speech with disfluencies can result in noisy Automatic Speech Recognition (ASR) transcripts, which affects downstream tasks like machine translation. In this paper, we propose an adversarially-trained sequence-tagging model for Disfluency Correction (DC) that utilizes a small amount of labeled real disfluent data in conjunction with a large amount of unlabeled data. We show the benefit of our proposed technique, which crucially depends on synthetically generated disfluent data, by evaluating it for DC in three Indian languages- Bengali, Hindi, and Marathi (all from the Indo-Aryan family). Our technique also performs well in removing stuttering disfluencies in ASR transcripts introduced by speech impairments. We achieve an average 6.15 points improvement in F1-score over competitive baselines across all three languages mentioned. To the best of our knowledge, we are the first to utilize adversarial training for DC and use it to correct stuttering disfluencies in English, establishing a new benchmark for this task.
PDF Accepted for Findings of ACL 2023
点此查看论文截图
OpenSR: Open-Modality Speech Recognition via Maintaining Multi-Modality Alignment
Authors:Xize Cheng, Tao Jin, Linjun Li, Wang Lin, Xinyu Duan, Zhou Zhao
Speech Recognition builds a bridge between the multimedia streaming (audio-only, visual-only or audio-visual) and the corresponding text transcription. However, when training the specific model of new domain, it often gets stuck in the lack of new-domain utterances, especially the labeled visual utterances. To break through this restriction, we attempt to achieve zero-shot modality transfer by maintaining the multi-modality alignment in phoneme space learned with unlabeled multimedia utterances in the high resource domain during the pre-training \cite{shi2022learning}, and propose a training system Open-modality Speech Recognition (\textbf{OpenSR}) that enables the models trained on a single modality (e.g., audio-only) applicable to more modalities (e.g., visual-only and audio-visual). Furthermore, we employ a cluster-based prompt tuning strategy to handle the domain shift for the scenarios with only common words in the new domain utterances. We demonstrate that OpenSR enables modality transfer from one to any in three different settings (zero-, few- and full-shot), and achieves highly competitive zero-shot performance compared to the existing few-shot and full-shot lip-reading methods. To the best of our knowledge, OpenSR achieves the state-of-the-art performance of word error rate in LRS2 on audio-visual speech recognition and lip-reading with 2.7\% and 25.0\%, respectively. The code and demo are available at https://github.com/Exgc/OpenSR.
PDF Accepted to ACL2023 (Oral)
点此查看论文截图
Modality Influence in Multimodal Machine Learning
Authors:Abdelhamid Haouhat, Slimane Bellaouar, Attia Nehar, Hadda Cherroun
Multimodal Machine Learning has emerged as a prominent research direction across various applications such as Sentiment Analysis, Emotion Recognition, Machine Translation, Hate Speech Recognition, and Movie Genre Classification. This approach has shown promising results by utilizing modern deep learning architectures. Despite the achievements made, challenges remain in data representation, alignment techniques, reasoning, generation, and quantification within multimodal learning. Additionally, assumptions about the dominant role of textual modality in decision-making have been made. However, limited investigations have been conducted on the influence of different modalities in Multimodal Machine Learning systems. This paper aims to address this gap by studying the impact of each modality on multimodal learning tasks. The research focuses on verifying presumptions and gaining insights into the usage of different modalities. The main contribution of this work is the proposal of a methodology to determine the effect of each modality on several Multimodal Machine Learning models and datasets from various tasks. Specifically, the study examines Multimodal Sentiment Analysis, Multimodal Emotion Recognition, Multimodal Hate Speech Recognition, and Multimodal Disease Detection. The study objectives include training SOTA MultiModal Machine Learning models with masked modalities to evaluate their impact on performance. Furthermore, the research aims to identify the most influential modality or set of modalities for each task and draw conclusions for diverse multimodal classification tasks. By undertaking these investigations, this research contributes to a better understanding of the role of individual modalities in multi-modal learning and provides valuable insights for future advancements in this field.
PDF 10 pages
点此查看论文截图
Audio-Visual Speech Enhancement With Selective Off-Screen Speech Extraction
Authors:Tomoya Yoshinaga, Keitaro Tanaka, Shigeo Morishima
This paper describes an audio-visual speech enhancement (AV-SE) method that estimates from noisy input audio a mixture of the speech of the speaker appearing in an input video (on-screen target speech) and of a selected speaker not appearing in the video (off-screen target speech). Although conventional AV-SE methods have suppressed all off-screen sounds, it is necessary to listen to a specific pre-known speaker’s speech (e.g., family member’s voice and announcements in stations) in future applications of AV-SE (e.g., hearing aids), even when users’ sight does not capture the speaker. To overcome this limitation, we extract a visual clue for the on-screen target speech from the input video and a voiceprint clue for the off-screen one from a pre-recorded speech of the speaker. Two clues from different domains are integrated as an audio-visual clue, and the proposed model directly estimates the target mixture. To improve the estimation accuracy, we introduce a temporal attention mechanism for the voiceprint clue and propose a training strategy called the muting strategy. Experimental results show that our method outperforms a baseline method that uses the state-of-the-art AV-SE and speaker extraction methods individually in terms of estimation accuracy and computational efficiency.
PDF Accepted by EUSIPCO 2023
点此查看论文截图
Computational Language Assessment: Open Brain AI
Authors:Charalambos Themistocleous
Language assessment plays a crucial role in diagnosing and treating individuals with speech, language, and communication disorders caused by neurogenic conditions, whether developmental or acquired. However, traditional manual assessment methods have several drawbacks. They are often laborious and time-consuming to administer and score, causing additional patient stress. Moreover, they divert valuable resources from treatment. To address these challenges, we introduce Open Brain AI (openbrainai.com), a computational platform that harnesses innovative AI techniques, including machine learning and natural language processing, to automatically analyze spoken and written speech productions. The platform leverages state-of-the-art AI techniques and aims to present a promising advancement in language assessment. Its ability to provide reliable and efficient measurements can enhance the accuracy of diagnoses and optimize treatment strategies for individuals with speech, language, and communication disorders. Furthermore, the automation and objectivity offered by the platform alleviate the burden on clinicians, enabling them to streamline their workflow and allocate more time and resources to direct patient care. Notably, the platform is freely accessible, empowering clinicians to conduct critical analyses of their data and allowing them to allocate more attention to other critical aspects of therapy and treatment.
PDF 1 figure, associated web application: openbrainai.com
点此查看论文截图
Multimodal Audio-textual Architecture for Robust Spoken Language Understanding
Authors:Anderson R. Avila, Mehdi Rezagholizadeh, Chao Xing
Recent voice assistants are usually based on the cascade spoken language understanding (SLU) solution, which consists of an automatic speech recognition (ASR) engine and a natural language understanding (NLU) system. Because such approach relies on the ASR output, it often suffers from the so-called ASR error propagation. In this work, we investigate impacts of this ASR error propagation on state-of-the-art NLU systems based on pre-trained language models (PLM), such as BERT and RoBERTa. Moreover, a multimodal language understanding (MLU) module is proposed to mitigate SLU performance degradation caused by errors present in the ASR transcript. The MLU benefits from self-supervised features learned from both audio and text modalities, specifically Wav2Vec for speech and Bert/RoBERTa for language. Our MLU combines an encoder network to embed the audio signal and a text encoder to process text transcripts followed by a late fusion layer to fuse audio and text logits. We found that the proposed MLU showed to be robust towards poor quality ASR transcripts, while the performance of BERT and RoBERTa are severely compromised. Our model is evaluated on five tasks from three SLU datasets and robustness is tested using ASR transcripts from three ASR engines. Results show that the proposed approach effectively mitigates the ASR error propagation problem, surpassing the PLM models’ performance across all datasets for the academic ASR engine.
PDF
点此查看论文截图
PauseSpeech: Natural Speech Synthesis via Pre-trained Language Model and Pause-based Prosody Modeling
Authors:Ji-Sang Hwang, Sang-Hoon Lee, Seong-Whan Lee
Although text-to-speech (TTS) systems have significantly improved, most TTS systems still have limitations in synthesizing speech with appropriate phrasing. For natural speech synthesis, it is important to synthesize the speech with a phrasing structure that groups words into phrases based on semantic information. In this paper, we propose PuaseSpeech, a speech synthesis system with a pre-trained language model and pause-based prosody modeling. First, we introduce a phrasing structure encoder that utilizes a context representation from the pre-trained language model. In the phrasing structure encoder, we extract a speaker-dependent syntactic representation from the context representation and then predict a pause sequence that separates the input text into phrases. Furthermore, we introduce a pause-based word encoder to model word-level prosody based on pause sequence. Experimental results show PauseSpeech outperforms previous models in terms of naturalness. Furthermore, in terms of objective evaluations, we can observe that our proposed methods help the model decrease the distance between ground-truth and synthesized speech. Audio samples are available at https://jisang93.github.io/pausespeech-demo/.
PDF 13 pages, 4 figures, 3 tables, under reivew
点此查看论文截图
Unsupervised speech enhancement with deep dynamical generative speech and noise models
Authors:Xiaoyu Lin, Simon Leglaive, Laurent Girin, Xavier Alameda-Pineda
This work builds on a previous work on unsupervised speech enhancement using a dynamical variational autoencoder (DVAE) as the clean speech model and non-negative matrix factorization (NMF) as the noise model. We propose to replace the NMF noise model with a deep dynamical generative model (DDGM) depending either on the DVAE latent variables, or on the noisy observations, or on both. This DDGM can be trained in three configurations: noise-agnostic, noise-dependent and noise adaptation after noise-dependent training. Experimental results show that the proposed method achieves competitive performance compared to state-of-the-art unsupervised speech enhancement methods, while the noise-dependent training configuration yields a much more time-efficient inference process.
PDF
点此查看论文截图
GEmo-CLAP: Gender-Attribute-Enhanced Contrastive Language-Audio Pretraining for Speech Emotion Recognition
Authors:Yu Pan, Yanni Hu, Yuguang Yang, Jixun Yao, Wen Fei, Lei Ma, Heng Lu
Contrastive Language-Audio Pretraining (CLAP) has recently exhibited impressive success in diverse fields. In this paper, we propose GEmo-CLAP, a kind of efficient gender-attribute-enhanced CLAP model for speech emotion recognition (SER). Specifically, we first build an effective emotion CLAP model termed Emo-CLAP for SER, utilizing various self-supervised learning based pre-trained models. Then, considering the importance of the gender attribute in speech emotion modeling, two GEmo-CLAP approaches are further proposed to integrate the emotion and gender information of speech signals, forming more reasonable objectives. Extensive experiments conducted on the IEMOCAP corpus demonstrate that our proposed two GEmo-CLAP approaches consistently outperform the baseline Emo-CLAP with different pre-trained models, while also achieving superior recognition performance compared with other state-of-the-art methods.
PDF 5 pages
点此查看论文截图