A key challenge in machine learning is to generalize from training data to an application domain of interest. This work extends the recently-proposed mixture invariant training (MixIT) algorithm to perform unsupervised learning in the multi-channel setting. We use MixIT to train a model on far-field microphone array recordings of overlapping reverberant and noisy speech from the AMI Corpus. The models are trained on both supervised and unsupervised training data, and are tested on real AMI recordings containing overlapping speech. To objectively evaluate our models, we also use a synthetic multi-channel AMI test set. Holding network architectures constant, we find that semi-supervised fine-tuning of a model pretrained on a large and diverse single-channel dataset yields the largest improvement to SI-SNR and to human listening ratings across synthetic and real datasets, outperforming supervised models trained on well-matched synthetic data. Our results demonstrate that unsupervised learning through MixIT enables model adaptation on both single-and multi-channel real-world speech recordings.

We propose the novel task of distance-based sound separation, where sounds are separated based only on their distance from a single microphone.In the context of assisted listening devices, proximity provides a simple criterion for sound selection in noisy environments that would allow the user to focus on sounds relevant to a local conversation.We demonstrate the feasibility of this approach by training a neural network to separate near sounds from far sounds in single channel synthetic reverberant mixtures, relative to a threshold distance defining the boundary between near and far.With a single nearby speaker and four distant speakers, the model improves scale-invariant signal to noise ratio by 4.4 dB for near sounds and 6.8 dB for far sounds.

We present an end-to-end deep network model that performs meeting diarization from single-channel audio recordings. End-to-end diarization models have the advantage of handling speaker overlap and enabling straightforward handling of discriminative training, unlike traditional clustering-based diarization methods. The proposed system is designed to handle meetings with unknown numbers of speakers, using variable-number permutation-invariant cross-entropy based loss functions. We introduce several components that appear to help with diarization performance, including a local convolutional network followed by a global self-attention module, multitask transfer learning using a speaker identification component, and a sequential approach where the model is refined with a second stage. These are trained and validated on simulated meeting data based on LibriSpeech and LibriTTS datasets; final evaluations are done using LibriCSS, which consists of simulated meetings recorded using real acoustics via loudspeaker playback. The proposed model performs better than previously proposed end-to-end diarization models on these data.

This work introduces sequential neural beamforming, which alternates between neural network based spectral separation and beamforming based spatial separation. Our neural networks for separation use an advanced convolutional architecture trained with a novel stabilized signal-to-noise ratio loss function. For beamforming, we explore multiple ways of computing time-varying covariance matrices, including factorizing the spatial covariance into a time-varying amplitude component and a time-invariant spatial component, as well as using block-based techniques. In addition, we introduce a multi-frame beamforming method which improves the results significantly by adding contextual frames to the beamforming formulations. We extensively evaluate and analyze the effects of window size, block size, and multi-frame context size for these methods. Our best method utilizes a sequence of three neural separation and multi-frame time-invariant spatial beamforming stages, and demonstrates an average improvement of 2.75 dB in scale-invariant signal-to-noise ratio and 14.2% absolute reduction in a comparative speech recognition metric across four challenging reverberant speech enhancement and separation tasks. We also use our three-speaker separation model to separate real recordings in the LibriCSS evaluation set into non-overlapping tracks, and achieve a better word error rate as compared to a baseline mask based beamformer.

We introduce VoiceFilter-Lite, a single-channel source separation model that runs on the device to preserve only the speech signals from a target user, as part of a streaming speech recognition system.Delivering such a model presents numerous challenges: It should improve the performance when the input signal consists of overlapped speech, and must not hurt the speech recognition performance under all other acoustic conditions.Besides, this model must be tiny, fast, and perform inference in a streaming fashion, in order to have minimal impact on CPU, memory, battery and latency.We propose novel techniques to meet these multi-faceted requirements, including using a new asymmetric loss, and adopting adaptive runtime suppression strength.We also show that such a model can be quantized as a 8-bit integer model and run in realtime.

In recent years, rapid progress has been made on the problem of single-channel sound separation using supervised training of deep neural networks. In such supervised approaches, a model is trained to predict the component sources from synthetic mixtures created by adding up isolated ground-truth sources. Reliance on this synthetic training data is problematic because good performance depends upon the degree of match between the training data and real-world audio, especially in terms of the acoustic conditions and distribution of sources. The acoustic properties can be challenging to accurately simulate, and the distribution of sound types may be hard to replicate. In this paper, we propose a completely unsupervised method, mixture invariant training (MixIT), that requires only single-channel acoustic mixtures. In MixIT, training examples are constructed by mixing together existing mixtures, and the model separates them into a variable number of latent sources, such that the separated sources can be remixed to approximate the original mixtures. We show that MixIT can achieve competitive performance compared to supervised methods on speech separation. Using MixIT in a semi-supervised learning setting enables unsupervised domain adaptation and learning from large amounts of real-world data without ground-truth source waveforms. In particular, we significantly improve reverberant speech separation performance by incorporating reverberant mixtures, train a speech enhancement system from noisy mixtures, and improve universal sound separation by incorporating a large amount of in-the-wild data.

In recent years, deep networks have led to dramatic improvements in speech enhancement by framing it as a data-driven pattern recognition problem. In many modern enhancement systems, large amounts of data are used to train a deep network to estimate masks for complex-valued short-time Fourier transforms (STFTs) to suppress noise and preserve speech. However, current masking approaches often neglect two important constraints: STFT consistency and mixture consistency. Without STFT consistency, the system's output is not necessarily the STFT of a time-domain signal, and without mixture consistency, the sum of the estimated sources does not necessarily equal the input mixture. Furthermore, the only previous approaches that apply mixture consistency use real-valued masks; mixture consistency has been ignored for complex-valued masks.

In this paper, we present a novel system that separates the voice of a target speaker from multi-speaker signals, by making use of a reference signal from the target speaker.We achieve this by training two separate neural networks: (1) A speaker recognition network that produces speaker-discriminative embeddings;(2) A spectrogram masking network that takes both noisy spectrogram and speaker embedding as input, and produces a mask.Our system significantly reduces the speech recognition WER on multi-speaker signals, with minimal WER degradation on single-speaker signals.

Speech activity detection (or endpointing) is an important processing step for applications such as speech recognition, language identification and speaker diarization.Both audio-and vision-based approaches have been used for this task in various settings, often tailored toward end applications.However, much of the prior work reports results in synthetic settings, on taskspecific datasets, or on datasets that are not openly available.This makes it difficult to compare approaches and understand their strengths and weaknesses.In this paper, we describe a new dataset which we will release publicly containing densely labeled speech activity in YouTube videos 1 , with the goal of creating a shared, available dataset for this task.The labels in the dataset annotate three different speech activity conditions: clean speech, speech co-occurring with music, and speech cooccurring with noise, which enable analysis of model performance in more challenging conditions based on the presence of overlapping noise.We report benchmark performance numbers on AVA-Speech using off-the-shelf, state-of-the-art audio and vision models that serve as a baseline to facilitate future research.

Convolutional Neural Networks (CNNs) have proven very effective in image classification and show promise for audio. We use various CNN architectures to classify the soundtracks of a dataset of 70M training videos (5.24 million hours) with 30,871 video-level labels. We examine fully connected Deep Neural Networks (DNNs), AlexNet [1], VGG [2], Inception [3], and ResNet [4]. We investigate varying the size of both training set and label vocabulary, finding that analogs of the CNNs used in image classification do well on our audio classification task, and larger training and label sets help up to a point. A model using embeddings from these classifiers does much better than raw features on the Audio Set [5] Acoustic Event Detection (AED) classification task.

Multichannel automatic speech recognition (ASR) systems commonly separate speech enhancement, including localization, beamforming, and postfiltering, from acoustic modeling. In this paper, we perform multichannel enhancement jointly with acoustic modeling in a deep neural network framework. Inspired by beamforming, which leverages differences in the fine time structure of the signal at different microphones to filter energy arriving from different directions, we explore modeling the raw time-domain waveform directly. We introduce a neural network architecture, which performs multichannel filtering in the first layer of the network, and show that this network learns to be robust to varying target speaker direction of arrival, performing as well as a model that is given oracle knowledge of the true target speaker direction. Next, we show how performance can be improved by factoring the first layer to separate the multichannel spatial filtering operation from a single channel filterbank which computes a frequency decomposition. We also introduce an adaptive variant, which updates the spatial filter coefficients at each time frame based on the previous inputs. Finally, we demonstrate that these approaches can be implemented more efficiently in the frequency domain. Overall, we find that such multichannel neural networks give a relative word error rate improvement of more than 5% compared to a traditional beamforming-based multichannel ASR system and more than 10% compared to a single channel waveform model.

Multichannel ASR systems commonly separate speech enhancement, including localization, beamforming and postfiltering, from acoustic modeling. Recently, we explored doing multichannel enhancement jointly with acoustic modeling, where beamforming and frequency decomposition was folded into one layer of the neural network [1, 2]. In this paper, we explore factoring these operations into separate layers in the network. Furthermore, we explore using multi-task learning (MTL) as a proxy for postfiltering, where we train the network to predict "clean" features as well as context-dependent states. We find that with the factored architecture, we can achieve a 10% relative improvement in WER over a single channel and a 5% relative improvement over the unfactored model from [1] on a 2,000-hour Voice Search task. In addition, by incorporating MTL, we can achieve 11% and 7% relative improvements over single channel and unfactored multichannel models, respectively.

Multichannel ASR systems commonly use separate modules to perform speech enhancement and acoustic modeling. In this paper, we present an algorithm to do multichannel enhancement jointly with the acoustic model, using a raw waveform convolutional LSTM deep neural network (CLDNN). We will show that our proposed method offers ~5% relative improvement in WER over a log-mel CLDNN trained on multiple channels. Analysis shows that the proposed network learns to be robust to varying angles of arrival for the target speaker, and performs as well as a model that is given oracle knowledge of the true location. Finally, we show that training such a network on inputs captured using multiple (linear) array configurations results in a model that is robust to a range of microphone spacings.

Standard deep neural network-based acoustic models for automatic speech recognition (ASR) rely on hand-engineered input features, typically log-mel filterbank magnitudes. In this paper, we describe a convolutional neural network - deep neural network (CNN-DNN) acoustic model which takes raw multichannel waveforms as input, i.e. without any preceding feature extraction, and learns a similar feature representation through supervised training. By operating directly in the time domain, the network is able to take advantage of the signal's fine time structure that is discarded when computing filterbank magnitude features. This structure is especially useful when analyzing multichannel inputs, where timing differences between input channels can be used to localize a signal in space. The first convolutional layer of the proposed model naturally learns a filterbank that is selective in both frequency and direction of arrival, i.e. a bank of bandpass beamformers with an auditory-like frequency scale. When trained on data corrupted with noise coming from different spatial locations, the network learns to filter them out by steering nulls in the directions corresponding to the noise sources. Experiments on a simulated multichannel dataset show that the proposed acoustic model outperforms a DNN that uses log-mel filterbank magnitude features under noisy and reverberant conditions.

Learning an acoustic model directly from the raw waveform has been an active area of research. However, waveform-based models have not yet matched the performance of log-mel trained neural networks. We will show that raw wave-form features match the performance of log-mel filterbank ener-gies when used with a state-of-the-art CLDNN acoustic model trained on over 2,000 hours of speech. Specifically, we will show the benefit of the CLDNN, namely the time convolution layer in reducing temporal variations, the frequency convolution layer for preserving locality and reducing frequency variations, as well as the LSTM layers for temporal modeling. In addition, by stacking raw waveform features with log-mel features, we achieve a 3 % relative reduction in word error rate. 1.

We present an algorithm to find a low-dimensional decomposition of a spectrogram by formulating this as a regularized non-negative matrix factorization (NMF) problem with a regularization term chosen to encourage independence. This algorithm provides a better decomposition than standard NMF when the underlying sources are independent. It is directly applicable to non-square matrices, and it makes better use of additional observation streams than previous nonnegative ICA algorithms.

We describe an algorithm that performs regularized non-negative matrix factorization (NMF) to find independent components in nonnegative data. Previous techniques proposed for this purpose require the data to be grounded, with support that goes down to 0 along each dimension. In our work, this requirement is eliminated. Based on it, we present a technique to find a low-dimensional decomposition of spectrograms by casting it as a problem of discovering independent non-negative components from it. The algorithm itself is implemented as regularized non-negative matrix factorization (NMF). Unlike other ICA algorithms, this algorithm computes the mixing matrix rather than an unmixing matrix. This algorithm provides a better decomposition than standard NMF when the underlying sources are independent. It makes better use of additional observation streams than previous nonnegative ICA algorithms. Index Terms — matrix decomposition, ICA 1.

We present a technique for denoising speech using nonnegative matrix factorization (NMF) in combination with statistical speech and noise models. We compare our new technique to standard NMF and to a state-of-the-art Wiener filter implementation and show improvements in speech quality across a range of interfering noise types.

We present a tecchnique for denoising speech using temporally regularized nonnegative matrix factorization (NMF). In previ-ous work [1], we used a regularized NMF update to impose structure within each audio frame. In this paper, we add frame-to-frame regularization across time and show that this additional regularization can also improve our speech denoising results. We evaluate our algorithm on a range of nonstationary noise types and outperform a state-of-the-art Wiener filter implemen-tation. Index Terms: speech enhancement, source separation, speech modeling, speech processing

We present a novel genre-independent SVM framework for detecting scene changes in broadcast video. Our framework works on content from a diverse range of genres by allowing sets of features, extracted from both audio and video streams, to be combined and compared automatically without the use of explicit thresholds. For ground truth, we use hand-labeled video scene boundaries from a wide variety of broadcast genres to generate positive and negative samples for the SVM. Our experiments include high-and low-level audio features such as semantic histograms and distances between Gaussian models, as well as video features such as shot cut positions. We evaluate the importance of these measures in a structured frame-work, with performance comparisons obtained via ROC curves. We achieve over 70% detection rate for 10% false positive rate on our corpus of over 7.5 hours of data collected from news, talk shows, sitcoms, dramas, music videos, and how-to shows.

Speech source localization in reverberant environments has proved difficult for automated microphone array systems. Because of its nonstationary nature, certain features observable in the reverberant speech signal, such as sudden increases in audio energy, provide cues to indicate time-frequency regions that are particularly useful for audio localization. We exploit these cues by learning a mapping from reverberated signal spectrograms to localization precision using ridge regression. Using the learned mappings in the generalized cross-correlation framework, we demonstrate improved localization performance. Additionally, the resulting mappings exhibit behavior consistent with the well-known precedence effect from psychoacoustic studies

Audio source localization in reverberant environments is difficult for automated microphone array systems. Certain features observable in the audio signal, such as sudden increases in audio energy, provide cues to indicate time-frequency regions that are particularly useful for audio localization, but previous approaches have not systematically exploited these cues. We learn a mapping from reverberated signal spectrograms to localization precision using ridge regression. The resulting mappings exhibit behavior consistent with the well-known precedence effect from psychoacoustic studies. Using the learned mappings, we demonstrate improved localization performance.

We present an approach to detecting and recognizing spoken isolated phrases based solely on visual input. We adopt an architecture that first employs discriminative detection of visual speech and articulate features, and then performs recognition using a model that accounts for the loose synchronization of the feature streams. Discriminative classifiers detect the subclass of lip appearance corresponding to the presence of speech, and further decompose it into features corresponding to the physical components of articulate production. These components often evolve in a semi-independent fashion, and conventional viseme-based approaches to recognition fail to capture the resulting co-articulation effects. We present a novel dynamic Bayesian network with a multi-stream structure and observations consisting of articulate feature classifier scores, which can model varying degrees of co-articulation in a principled way. We evaluate our visual-only recognition system on a command utterance task. We show comparative results on lip detection and speech/non-speech classification, as well as recognition performance against several baseline systems

In this paper, we present a system that combines sound and vision to track multiple people. In a cluttered or noisy scene, multi-person tracking estimates have a distinctly non-Gaussian distribution. We apply a particle filter with audio and video state components, and derive observation likelihood methods based on both audio and video measurements. Our state includes the number of people present, their positions, and whether each person is talking. We show experiments in an environment with sparse microphones and monocular cameras. Our results show that our system can accurately track the locations and speech activity of a varying number of people.

We demonstrate an audio-visual tracking system for meeting analysis. A stereo camera and a microphone array are used to track multiple people and their speech activity in real-time. Our system can estimate the location of multiple people, detect the current speaker and build a model of interaction between people in a meeting.

This paper presents a multi-modal approach to locate a speaker in a scene and determine to whom he or she is speaking. We present a simple probabilistic framework that combines multiple cues derived from both audio and video information. A purely visual cue is obtained using a head tracker to identify possible speakers in a scene and provide both their 3-D positions and orientation. In addition, estimates of the audio signal's direction of arrival are obtained with the help of a two-element microphone array. A third cue measures the association between the audio and the tracked regions in the video. Integrating these cues provides a more robust solution than using any single cue alone. The usefulness of our approach is shown in our results for video sequences with two or more people in a prototype interactive kiosk environment.

Steerable microphone arrays provide a flexible infrastructure for audio source separation. In order for them to be used effectively in intelligent environments, there must be a mechanism in place for steering the focus of the array to the sound source. Audio-only steering techniques often perform poorly in the presence of multiple sound sources or strong reverberation. Video-only techniques can achieve high spatial precision but require that the audio and video subsystems be accurately calibrated to preserve this precision. We present an audio-video localization technique that combines the benefits of the two modalities. We implement our technique in a test environment containing multiple stereo cameras and a room-sized microphone array. Our technique achieves an 8.9 dB improvement over a single far-field microphone, a 6.7 dB improvement over source separation based on video-only localization, and a 0.3 dB improvement over separation based on audio-only localization.

When faced with a distant speaker at a known location in a noisy environment, a microphone array can provide a significantly improved audio signal for speech recognition. Estimating the location of a speaker in a reverberant environment from audio information alone can be quite difficult, so we use an array of video cameras to aid localization. Stereo processing techniques are used on pairs of cameras, and foreground 3-D points are grouped to estimate the trajectory of people as they move in an environment. These trajectories are used to guide a microphone array beamformer. Initial results using this system for speech recognition demonstrate increased recognition rates compared to non-array processing techniques.