CTC compressor can be an effective approach to integrate audio encoders to decoder-only models, which has gained growing interest for different speech applications. In this work, we propose a novel CTC compressor based joint speech and text training (CJST) framework for decoder-only ASR. CJST matches speech and text modalities from both directions by exploring a simple modality adaptor and several features of the CTC compressor, including sequence compression, on-the-fly forced peaky alignment and CTC class embeddings. Experimental results on the Librispeech and TED-LIUM2 corpora show that the proposed CJST achieves an effective text injection without the need of duration handling, leading to the best performance for both in-domain and cross-domain scenarios. We also provide a comprehensive study on CTC compressor, covering various compression modes, edge case handling and behavior under both clean and noisy data conditions, which reveals the most robust setting to use CTC compressor for decoder-only models.

We introduce Ego4D, a massive-scale egocentric video dataset and benchmark suite. It offers 3,670 hours of daily-life activity video spanning hundreds of scenarios (household, outdoor, workplace, leisure, etc.) captured by 931 unique camera wearers from 74 worldwide locations and 9 different countries. The approach to collection is designed to uphold rigorous privacy and ethics standards, with consenting participants and robust de-identification procedures where relevant. Ego4D dramatically expands the volume of diverse egocentric video footage publicly available to the research community. Portions of the video are accompanied by audio, 3D meshes of the environment, eye gaze, stereo, and/or synchronized videos from multiple egocentric cameras at the same event. Furthermore, we present a host of new benchmark challenges centered around understanding the first-person visual experience in the past (querying an episodic memory), present (analyzing hand-object manipulation, audio-visual conversation, and social interactions), and future (forecasting activities). By publicly sharing this massive annotated dataset and benchmark suite, we aim to push the frontier of first-person perception.

Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language. In this work, we propose the use of an adaptive masking approach in two scenarios for pruning a multilingual ASR model efficiently, each resulting in sparse monolingual models or a sparse multilingual model (named as Dynamic ASR Pathways). Our approach dynamically adapts the subnetwork, avoiding premature decisions about a fixed sub-network structure. We show that our approach outperforms existing pruning methods when targeting sparse monolingual models. Further, we illustrate that Dynamic ASR Pathways jointly discovers and trains better sub-networks (pathways) of a single multilingual model by adapting from different sub-network initializations, thereby reducing the need for language-specific pruning.

A singing voice conversion model converts a song in the voice of an arbitrary source singer to the voice of a target singer. Recently, methods that leverage self-supervised audio representations such as HuBERT and Wav2Vec 2.0 have helped further the state-of-the-art. Though these methods produce more natural and melodic singing outputs, they often rely on confusion and disentanglement losses to render the self-supervised representations speaker and pitch-invariant. In this paper, we circumvent disentanglement training and propose a new model that leverages ASR fine-tuned self-supervised representations as inputs to a HiFi-GAN neural vocoder for singing voice conversion. We experiment with different f0encoding schemes and show that an f0harmonic generation module that uses a parallel bank of transposed convolutions (PBTC) alongside ASR fine-tuned Wav2Vec 2.0 features results in the best singing voice conversion quality. Additionally, the model is capable of making a spoken voice sing. We also show that a simple f0shifting scheme during inference helps retain singer identity and bolsters the performance of our singing voice conversion model. Our results are backed up by extensive MOS studies that compare different ablations and baselines.

Large-scale generative models such as GPT and DALL-E have revolutionized the research community. These models not only generate high fidelity outputs, but are also generalists which can solve tasks not explicitly taught. In contrast, speech generative models are still primitive in terms of scale and task generalization. In this paper, we present Voicebox, the most versatile text-guided generative model for speech at scale. Voicebox is a non-autoregressive flow-matching model trained to infill speech, given audio context and text, trained on over 50K hours of speech that are not filtered or enhanced. Similar to GPT, Voicebox can perform many different tasks through in-context learning, but is more flexible as it can also condition on future context. Voicebox can be used for mono or cross-lingual zero-shot text-to-speech synthesis, noise removal, content editing, style conversion, and diverse sample generation. In particular, Voicebox outperforms the state-of-the-art zero-shot TTS model VALL-E on both intelligibility (5.9\% vs 1.9\% word error rates) and audio similarity (0.580 vs 0.681) while being up to 20 times faster. Audio samples can be found in \url{https://voicebox.metademolab.com}.

We introduce Ego4D, a massive-scale egocentric video dataset and benchmark suite. It offers 3,670 hours of dailylife activity video spanning hundreds of scenarios (household, outdoor, workplace, leisure, etc.) captured by 931 unique camera wearers from 74 worldwide locations and 9 different countries. The approach to collection is designed to uphold rigorous privacy and ethics standards, with consenting participants and robust de-identification procedures where relevant. Ego4D dramatically expands the volume of diverse egocentric video footage publicly available to the research community. Portions of the video are accompanied by audio, 3D meshes of the environment, eye gaze, stereo, and/or synchronized videos from multiple egocentric cameras at the same event. Furthermore, we present a host of new benchmark challenges centered around understanding the first-person visual experience in the past (querying an episodic memory), present (analyzing hand-object manipulation, audio-visual conversation, and social interactions), and future (forecasting activities). By publicly sharing this massive annotated dataset and benchmark suite, we aim to push the frontier of first-person perception. Project page: https://ego4d-data.org/

The problem of machine learning systems demonstrating bias towards specific groups of individuals has been studied extensively, particularly in the Facial Recognition area, but much less so in Automatic Speech Recognition (ASR). This paper presents initial Speech Recognition results on “Casual Conversations” – a publicly released 846 hour corpus designed to help researchers evaluate their computer vision and audio models for accuracy across a diverse set of metadata, including age, gender, and skin tone. The entire corpus has been manually transcribed, allowing for detailed ASR evaluations across these metadata. Multiple ASR models are evaluated, including models trained on LibriSpeech, 14,000 hour transcribed, and over 2 million hour untranscribed social media videos. Significant differences in word error rate across gender and skin tone are observed at times for all models. We are releasing human transcripts from the Casual Conversations dataset to encourage the community to develop a variety of techniques to reduce these statistical biases.

Concerns have been raised regarding performance disparity in automatic speech recognition (ASR) systems as they provide unequal transcription accuracy for different user groups defined by different attributes that include gender, dialect, and race. In this paper, we propose “equal accuracy ratio”, a novel inclusiveness measure for ASR systems that can be seamlessly integrated into the standard connectionist temporal classification (CTC) training pipeline of an end-to-end neural speech recognizer to increase the recognizer’s inclusiveness. We also create a novel multi-dialect benchmark dataset to study the inclusiveness of ASR, by combining data from existing corpora in seven dialects of English (African American, General American, Latino English, British English, Indian English, Afrikaaner English, and Xhosa English). Experiments on this multi-dialect corpus show that using the equal accuracy ratio as a regularization term along with CTC loss, succeeds in lowering the accuracy gap between user groups and reduces the recognition error rate compared with a non-regularized baseline. Experiments on additional speech corpora that have different user groups also confirm our findings.

Although speaker verification has conventionally been an audio-only task, some practical applications provide both audio and visual streams of input. In these cases, the visual stream provides complementary information and can often be leveraged in conjunction with the acoustics of speech to improve verification performance. In this study, we explore audio-visual approaches to speaker verification, starting with standard fusion techniques to learn joint audio-visual (AV) embeddings, and then propose a novel approach to handle cross-modal verification at test time. Specifically, we investigate unimodal and concatenation based AV fusion and report the lowest AV equal error rate (EER) of 0.7% on the VoxCeleb1 dataset using our best system. As these methods lack the ability to do cross-modal verification, we introduce a multi-view model which uses a shared classifier to map audio and video into the same space. This new approach achieves 28% EER on VoxCeleb1 in the challenging testing condition of cross-modal verification.

Kiran Ramnath, Leda Sari, Mark Hasegawa-Johnson, Chang Yoo. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2021.

Speaker adaptation and speaker change detection have both been studied extensively to improve automatic speech recognition (ASR). In many cases, these two problems are investigated separately: speaker change detection is implemented first to obtain single-speaker regions, and speaker adaptation is then performed using the derived speaker segments for improved ASR. However, in an online setting, we want to achieve both goals in a single pass. In this study, we propose a neural network architecture that learns a speaker embedding from which it can perform both speaker adaptation for ASR and speaker change detection. The proposed speaker embedding is computed using self-attention based on an auxiliary network attached to a main ASR network. ASR adaptation is then performed by subtracting, from the main network activations, a segment dependent affine transformation of the learned speaker embedding. In experiments on a broadcast news dataset and the Switchboard conversational dataset, we test our system on utterances with a change point in them and show that the proposed method achieves significantly better performance as compared to the unadapted main network (10-14% relative reduction in word error rate (WER)). The proposed architecture also outperforms three different speaker segmentation methods followed by ASR (around 10% relative reduction in WER).

Widely used automatic speech recognition (ASR) systems have been empirically demonstrated in various studies to be unfair, having higher error rates for some groups of users than others. One way to define fairness in ASR is to require that changing the demographic group affiliation of any individual (e.g., changing their gender, age, education or race) should not change the probability distribution across possible speech-to-text transcriptions. In the paradigm of counterfactual fairness, all variables independent of group affiliation (e.g., the text being read by the speaker) remain unchanged, while variables dependent on group affiliation (e.g., the speaker's voice) are counterfactually modified. Hence, we approach the fairness of ASR by training the ASR to minimize change in its outcome probabilities despite a counterfactual change in the individual's demographic attributes. Starting from the individualized counterfactual equal odds criterion, we provide relaxations to it and compare their performances for connectionist temporal classification (CTC) based end-to-end ASR systems. We perform our experiments on the Corpus of Regional African American Languages (CORAAL) and the LibriSpeech dataset to accommodate for differences due to gender, age, education, and race. We show that with counterfactual training, we can reduce average character error rates while achieving lower performance gap between demographic groups, and lower error standard deviation among individuals.

End-to-end spoken language understanding (SLU) systems are typically trained on large amounts of data. In many practical scenarios, the amount of labeled speech is often limited as opposed to text. In this study, we investigate the use of non-parallel speech and text to improve the performance of dialog act recognition as an example SLU task. We propose a multiview architecture that can handle each modality separately. To effectively train on such data, this model enforces the internal speech and text encodings to be similar using a shared classifier. On the Switchboard Dialog Act corpus, we show that pretraining the classifier using large amounts of text helps learning better speech encodings, resulting in up to 40% relatively higher classification accuracies. We also show that when the speech embeddings from an automatic speech recognition (ASR) system are used in this framework, the speech-only accuracy exceeds the performance of ASR-text based tests up to 15% relative and approaches the performance of using true transcripts.

We propose an unsupervised speaker adaptation method inspired by the neural Turing machine for end-to-end (E2E) automatic speech recognition (ASR). The proposed model contains a memory block that holds speaker i-vectors extracted from the training data and reads relevant i-vectors from the memory through an attention mechanism. The resulting memory vector (M-vector) is concatenated to the acoustic features or to the hidden layer activations of an E2E neural network model. The E2E ASR system is based on the joint connectionist temporal classification and attention-based encoder-decoder architecture. M-vector and i-vector results are compared for inserting them at different layers of the encoder neural network using the WSJ and TED-LIUM2 ASR benchmarks. We show that M-vectors, which do not require an auxiliary speaker embedding extraction system at test time, achieve similar word error rates (WERs) compared to i-vectors for single speaker utterances and significantly lower WERs for utterances in which there are speaker changes.

Spoken language understanding (SLU) datasets, like many other machine learning datasets, usually suffer from the label imbalance problem. Label imbalance usually causes the learned model to replicate similar biases at the output which raises the issue of unfairness to the minority classes in the dataset. In this work, we approach the fairness problem by maximizing the F-measure instead of accuracy in neural network model training. We propose a differentiable approximation to the F-measure and train the network with this objective using standard backpropagation. We perform experiments on two standard fairness datasets, Adult, and Communities and Crime, and also on speech-to-intent detection on the ATIS dataset and speech-to-image concept classification on the Speech-COCO dataset. In all four of these tasks, F-measure maximization results in improved micro-F1 scores, with absolute improvements of up to 8% absolute, as compared to models trained with the cross-entropy loss function. In the two multi-class SLU tasks, the proposed approach significantly improves class coverage, i.e., the number of classes with positive recall.

In scenarios where multiple speakers talk at the same time, it is important to be able to identify the talkers accurately.This paper presents an end-to-end system that integrates speech source extraction and speaker identification, and proposes a new way to jointly optimize these two parts by max-pooling the speaker predictions along the channel dimension.Residual attention permits us to learn spectrogram masks that are optimized for the purpose of speaker identification, while residual forward connections permit dilated convolution with a sufficiently large context window to guarantee correct streaming across syllable boundaries.End-to-end training results in a system that recognizes one speaker in a two-speaker broadcast speech mixture with 99.9% accuracy and both speakers with 93.9% accuracy, and that recognizes all speakers in three-speaker scenarios with 81.2% accuracy.

In this work, we investigate pre-training of neural network based speaker embeddings for low-latency speaker change detection. Our proposed system takes two speech segments, generates embeddings using shared Siamese layers and then classifies the concatenated embeddings depending on whether they are spoken by the same speaker. We investigate gender classification, contrastive loss and triplet loss based pre-training of the embedding layers and also joint training of the embedding layers along with a same/different classifier. Training is performed on 2-second single speaker segments based on ground truth speaker segmentation of broadcast news data. However, during test, we use the detection system in a practical low-latency setting for annotating automatic closed captions. In contrast to training, test pairs are now created around automatic speech recognition (ASR) based segmentation boundaries. The ASR segments are often shorter than 2 seconds causing duration mismatch during testing. In our experiments, although the baseline i-vector based classifier performs well, the proposed triplet loss based pre-training followed by joint training provides 7-50% relative F-measure improvement in matched and mismatched conditions. In addition, the degradation in performance is less severe for network based embeddings as compared to using i-vectors in the variable duration test conditions.

In this paper, we address the problem of defect detection in textile images, and present a novel hybrid method where independent vector analysis, a statistical method, is combined with wavelet transformation, a spectral method. Independent vector analysis, a generalization of independent component analysis, uses vectorized signals, thus, enables exploiting multiple datasets and offers a fully multivariate analysis. In this study, the multiple datasets are generated by wavelet transforming the texture image blocks of a predetermined size, and consequently, sub bands generated provide the dependent multiple datasets which are jointly processed to extract information from the dependencies present among them. Furthermore, subject diversity is introduced by taking these image blocks from multiple images corresponding to different textures in the TILDA database during the training phase. From this viewpoint, the proposed method may also be interpreted as a combination of independent vector analysis and group independent component analysis models. The proposed independent vector analysis based method is compared with the independent component analysis based method and some improvement from performance point of view is observed. Considering the results obtained, the proposed method can be an alternative solution to the defect detection problem.