In the field of spoken language understanding, systems like Whisper and Multilingual Massive Speech (MMS) have shown state-of-the-art performances. This study is dedicated to a comprehensive exploration of the Whisper and MMS systems, with a focus on assessing biases in automatic speech recognition (ASR) inherent to casual conversation speech specific to the Portuguese language. Our investigation encompasses various categories, including gender, age, skin tone color, and geo-location. Alongside traditional ASR evaluation metrics such as Word Error Rate (WER), we have incorporated p-value statistical significance for gender bias analysis. Furthermore, we extensively examine the impact of data distribution and empirically show that oversampling techniques alleviate such stereotypical biases. This research represents a pioneering effort in quantifying biases in the Portuguese language context through the application of MMS and Whisper, contributing to a better understanding of ASR systems' performance in multilingual settings.
Cued Speech (CS) is a pure visual coding method used by hearing-impaired people that combines lip reading with several specific hand shapes to make the spoken language visible. Automatic CS recognition (ACSR) seeks to transcribe visual cues of speech into text, which can help hearing-impaired people to communicate effectively. The visual information of CS contains lip reading and hand cueing, thus the fusion of them plays an important role in ACSR. However, most previous fusion methods struggle to capture the global dependency present in long sequence inputs of multi-modal CS data. As a result, these methods generally fail to learn the effective cross-modal relationships that contribute to the fusion. Recently, attention-based transformers have been a prevalent idea for capturing the global dependency over the long sequence in multi-modal fusion, but existing multi-modal fusion transformers suffer from both poor recognition accuracy and inefficient computation for the ACSR task. To address these problems, we develop a novel computation and parameter efficient multi-modal fusion transformer by proposing a novel Token-Importance-Aware Attention mechanism (TIAA), where a token utilization rate (TUR) is formulated to select the important tokens from the multi-modal streams. More precisely, TIAA firstly models the modality-specific fine-grained temporal dependencies over all tokens of each modality, and then learns the efficient cross-modal interaction for the modality-shared coarse-grained temporal dependencies over the important tokens of different modalities. Besides, a light-weight gated hidden projection is designed to control the feature flows of TIAA. The resulting model, named Economical Cued Speech Fusion Transformer (EcoCued), achieves state-of-the-art performance on all existing CS datasets, compared with existing transformer-based fusion methods and ACSR fusion methods.
In this work, we propose a novel cross-talk rejection framework for a multi-channel multi-talker setup for a live multiparty interactive show. Our far-field audio setup is required to be hands-free during live interaction and comprises four adjacent talkers with directional microphones in the same space. Such setups often introduce heavy cross-talk between channels, resulting in reduced automatic speech recognition (ASR) and natural language understanding (NLU) performance. To address this problem, we propose voice activity detection (VAD) model for all talkers using multichannel information, which is then used to filter audio for downstream tasks. We adopt a synthetic training data generation approach through playback and re-recording for such scenarios, simulating challenging speech overlap conditions. We train our models on this synthetic data and demonstrate that our approach outperforms single-channel VAD models and energy-based multi-channel VAD algorithm in various acoustic environments. In addition to VAD results, we also present multiparty ASR evaluation results to highlight the impact of using our VAD model for filtering audio in downstream tasks by significantly reducing the insertion error.
Social robots aim to establish long-term bonds with humans through engaging conversation. However, traditional conversational approaches, reliant on scripted interactions, often fall short in maintaining engaging conversations. This paper addresses this limitation by integrating large language models (LLMs) into social robots to achieve more dynamic and expressive conversations. We introduce a fully-automated conversation system that leverages LLMs to generate robot responses with expressive behaviors, congruent with the robot's personality. We incorporate robot behavior with two modalities: 1) a text-to-speech (TTS) engine capable of various delivery styles, and 2) a library of physical actions for the robot. We develop a custom, state-of-the-art emotion recognition model to dynamically select the robot's tone of voice and utilize emojis from LLM output as cues for generating robot actions. A demo of our system is available here. To illuminate design and implementation issues, we conduct a pilot study where volunteers chat with a social robot using our proposed system, and we analyze their feedback, conducting a rigorous error analysis of chat transcripts. Feedback was overwhelmingly positive, with participants commenting on the robot's empathy, helpfulness, naturalness, and entertainment. Most negative feedback was due to automatic speech recognition (ASR) errors which had limited impact on conversations. However, we observed a small class of errors, such as the LLM repeating itself or hallucinating fictitious information and human responses, that have the potential to derail conversations, raising important issues for LLM application.
While recent work shows promising results in expanding the capabilities of large language models (LLM) to directly understand and synthesize speech, an LLM-based strategy for modeling spoken dialogs remains elusive and calls for further investigation. This work proposes an extensive speech-text LLM framework, named the Unified Spoken Dialog Model (USDM), to generate coherent spoken responses with organic prosodic features relevant to the given input speech without relying on automatic speech recognition (ASR) or text-to-speech (TTS) solutions. Our approach employs a multi-step speech-text inference scheme that leverages chain-of-reasoning capabilities exhibited by the underlying LLM. We also propose a generalized speech-text pretraining scheme that helps with capturing cross-modal semantics. Automatic and human evaluations show that the proposed approach is effective in generating natural-sounding spoken responses, outperforming both prior and cascaded baselines. Detailed comparative studies reveal that, despite the cascaded approach being stronger in individual components, the joint speech-text modeling improves robustness against recognition errors and speech quality. Demo is available at https://unifiedsdm.github.io.
Error correction techniques have been used to refine the output sentences from automatic speech recognition (ASR) models and achieve a lower word error rate (WER). Previous works usually adopt end-to-end models and has strong dependency on Pseudo Paired Data and Original Paired Data. But when only pre-training on Pseudo Paired Data, previous models have negative effect on correction. While fine-tuning on Original Paired Data, the source side data must be transcribed by a well-trained ASR model, which takes a lot of time and not universal. In this paper, we propose UCorrect, an unsupervised Detector-Generator-Selector framework for ASR Error Correction. UCorrect has no dependency on the training data mentioned before. The whole procedure is first to detect whether the character is erroneous, then to generate some candidate characters and finally to select the most confident one to replace the error character. Experiments on the public AISHELL-1 dataset and WenetSpeech dataset show the effectiveness of UCorrect for ASR error correction: 1) it achieves significant WER reduction, achieves 6.83\% even without fine-tuning and 14.29\% after fine-tuning; 2) it outperforms the popular NAR correction models by a large margin with a competitive low latency; and 3) it is an universal method, as it reduces all WERs of the ASR model with different decoding strategies and reduces all WERs of ASR models trained on different scale datasets.
A novel feature, based on the chirp z-transform, that offers an improved representation of the underlying true spectrum is proposed. This feature, the chirp MFCC, is derived by computing the Mel frequency cepstral coefficients from the chirp magnitude spectrum, instead of the Fourier transform magnitude spectrum. The theoretical foundations for the proposal, and the experimental validation using product of likelihood Gaussians, to show the improved class separation offered by the proposed chirp MFCC, when compared with vanilla MFCC are discussed. Further, real world evaluation of the feature is performed using three diverse tasks, namely, speech-music classification, speaker identification, and speech commands recognition. It is shown in all three tasks that the proposed chirp MFCC offers considerable improvements.
Pre-trained Models (PTMs) have facilitated substantial progress in the field of Speech Emotion Recognition (SER). SER is an area with applications ranging from HumanComputer Interaction to Healthcare. Recent studies have leveraged various PTM representations as input features for downstream models for SER. PTM specifically pre-trained for paralinguistic tasks have obtained state-of-the-art (SOTA) performance for SER. However, such PTM haven't been evaluated for SER in multilingual settings and experimented only with English. So, we fill this gap, by performing a comprehensive comparative study of five PTMs (TRILLsson, wav2vec2, XLS-R, x-vector, Whisper) for assessing the effectiveness of paralingual PTM (TRILLsson) for SER across multiple languages. Representations from TRILLsson achieved the best performance among all the PTMs. This demonstrates that TRILLsson is able to effectively capture the various paralinguistic features from speech data for better SER. We also show that downstream models using TRILLsson representations achieve SOTA performance in terms of accuracy across various multi-lingual datasets.
\textbf{Objectives}: We aimed to investigate how errors from automatic speech recognition (ASR) systems affect dementia classification accuracy, specifically in the ``Cookie Theft'' picture description task. We aimed to assess whether imperfect ASR-generated transcripts could provide valuable information for distinguishing between language samples from cognitively healthy individuals and those with Alzheimer's disease (AD). \textbf{Methods}: We conducted experiments using various ASR models, refining their transcripts with post-editing techniques. Both these imperfect ASR transcripts and manually transcribed ones were used as inputs for the downstream dementia classification. We conducted comprehensive error analysis to compare model performance and assess ASR-generated transcript effectiveness in dementia classification. \textbf{Results}: Imperfect ASR-generated transcripts surprisingly outperformed manual transcription for distinguishing between individuals with AD and those without in the ``Cookie Theft'' task. These ASR-based models surpassed the previous state-of-the-art approach, indicating that ASR errors may contain valuable cues related to dementia. The synergy between ASR and classification models improved overall accuracy in dementia classification. \textbf{Conclusion}: Imperfect ASR transcripts effectively capture linguistic anomalies linked to dementia, improving accuracy in classification tasks. This synergy between ASR and classification models underscores ASR's potential as a valuable tool in assessing cognitive impairment and related clinical applications.
In this paper, we propose an efficient and accurate streaming speech recognition model based on the FastConformer architecture. We adapted the FastConformer architecture for streaming applications through: (1) constraining both the look-ahead and past contexts in the encoder, and (2) introducing an activation caching mechanism to enable the non-autoregressive encoder to operate autoregressively during inference. The proposed model is thoughtfully designed in a way to eliminate the accuracy disparity between the train and inference time which is common for many streaming models. Furthermore, our proposed encoder works with various decoder configurations including Connectionist Temporal Classification (CTC) and RNN-Transducer (RNNT) decoders. Additionally, we introduced a hybrid CTC/RNNT architecture which utilizes a shared encoder with both a CTC and RNNT decoder to boost the accuracy and save computation. We evaluate the proposed model on LibriSpeech dataset and a multi-domain large scale dataset and demonstrate that it can achieve better accuracy with lower latency and inference time compared to a conventional buffered streaming model baseline. We also showed that training a model with multiple latencies can achieve better accuracy than single latency models while it enables us to support multiple latencies with a single model. Our experiments also showed the hybrid architecture would not only speedup the convergence of the CTC decoder but also improves the accuracy of streaming models compared to single decoder models.