Audio-visual speaker tracking aims to determine the location of human targets in a scene using signals captured by a multi-sensor platform, whose accuracy and robustness can be improved by multi-modal fusion methods. Recently, several fusion methods have been proposed to model the correlation in multiple modalities. However, for the speaker tracking problem, the cross-modal interaction between audio and visual signals hasn't been well exploited. To this end, we present a novel Speaker Tracking Network (STNet) with a deep audio-visual fusion model in this work. We design a visual-guided acoustic measurement method to fuse heterogeneous cues in a unified localization space, which employs visual observations via a camera model to construct the enhanced acoustic map. For feature fusion, a cross-modal attention module is adopted to jointly model multi-modal contexts and interactions. The correlated information between audio and visual features is further interacted in the fusion model. Moreover, the STNet-based tracker is applied to multi-speaker cases by a quality-aware module, which evaluates the reliability of multi-modal observations to achieve robust tracking in complex scenarios. Experiments on the AV16.3 and CAV3D datasets show that the proposed STNet-based tracker outperforms uni-modal methods and state-of-the-art audio-visual speaker trackers.

Extracting direct-path spatial feature is crucial for sound source localization in adverse acoustic environments. This paper proposes IPDnet, a neural network that estimates direct-path inter-channel phase difference (DP-IPD) of sound sources from microphone array signals. The estimated DP-IPD can be easily translated to source location based on the known microphone array geometry. First, a full-band and narrow-band fusion network is adopted for DP-IPD estimation, in which combined narrow-band and full-band layers are responsible for estimating the raw DP-IPD information in one frequency band and capturing the frequency correlations of DP-IPD, respectively. Second, a new multi-track DP-IPD learning target is proposed for the localization of a flexible number of sound sources. Third, the network is extended to handle variable microphone arrays. This version of IPDnet is trained with a large set of different microphone arrays, and then it is able to infer the source locations using new microphone arrays not seen at training time. Experiments with multiple number of moving speakers are conducted on both simulated and real-world data, which show that the full-band and narrow-band fusion network and the proposed multi-track DP-IPD learning target together achieve excellent sound source localization performance. Moreover, the proposed variable-array model generalizes well to unseen microphone arrays.

Supervised learning methods have shown effectiveness in estimating spatial acoustic parameters such as time difference of arrival, direct-to-reverberant ratio and reverberation time. However, they still suffer from the simulation-to-reality generalization problem due to the mismatch between simulated and real-world acoustic characteristics and the deficiency of annotated real-world data. To this end, this work proposes a self-supervised method that takes full advantage of unlabeled data for spatial acoustic parameter estimation. First, a new pretext task, i.e. cross-channel signal reconstruction (CCSR), is designed to learn a universal spatial acoustic representation from unlabeled multi-channel microphone signals. We mask partial signals of one channel and ask the model to reconstruct them, which makes it possible to learn spatial acoustic information from unmasked signals and extract source information from the other microphone channel. An encoder-decoder structure is used to disentangle the two kinds of information. By fine-tuning the pre-trained spatial encoder with a small annotated dataset, this encoder can be used to estimate spatial acoustic parameters. Second, a novel multi-channel audio Conformer (MC-Conformer) is adopted as the encoder model architecture, which is suitable for both the pretext and downstream tasks. It is carefully designed to be able to capture the local and global characteristics of spatial acoustics exhibited in the time-frequency domain. Experimental results of five acoustic parameter estimation tasks on both simulated and real-world data show the effectiveness of the proposed method. To the best of our knowledge, this is the first self-supervised learning method in the field of spatial acoustic representation learning and multi-channel audio signal processing.

Multiple moving sound source localization in real-world scenarios remains a challenging issue due to interaction between sources, time-varying trajectories, distorted spatial cues, etc. In this work, we propose to use deep learning techniques to learn competing and time-varying direct-path phase differences for localizing multiple moving sound sources. A causal convolutional recurrent neural network is designed to extract the direct-path phase difference sequence from signals of each microphone pair. To avoid the assignment ambiguity and the problem of uncertain output-dimension encountered when simultaneously predicting multiple targets, the learning target is designed in a weighted sum format, which encodes source activity in the weight and direct-path phase differences in the summed value. The learned direct-path phase differences for all microphone pairs can be directly used to construct the spatial spectrum according to the formulation of steered response power (SRP). This deep neural network (DNN) based SRP method is referred to as SRP-DNN. The locations of sources are estimated by iteratively detecting and removing the dominant source from the spatial spectrum, in which way the interaction between sources is reduced. Experimental results on both simulated and real-world data show the superiority of the proposed method in the presence of noise and reverberation.

Graph convolutional networks have been widely used for skeleton-based action recognition due to their excellent modeling ability of non-Euclidean data. As the graph convolution is a local operation, it can only utilize the short-range joint dependencies and short-term trajectory but fails to directly model the distant joints relations and long-range temporal information that are vital to distinguishing various actions. To solve this problem, we present a multi-scale spatial graph convolution (MS-GC) module and a multi-scale temporal graph convolution (MT-GC) module to enrich the receptive field of the model in spatial and temporal dimensions. Concretely, the MS-GC and MT-GC modules decompose the corresponding local graph convolution into a set of sub-graph convolution, forming a hierarchical residual architecture. Without introducing additional parameters, the features will be processed with a series of sub-graph convolutions, and each node could complete multiple spatial and temporal aggregations with its neighborhoods. The final equivalent receptive field is accordingly enlarged, which is capable of capturing both short- and long-range dependencies in spatial and temporal domains. By coupling these two modules as a basic block, we further propose a multi-scale spatial temporal graph convolutional network (MST-GCN), which stacks multiple blocks to learn effective motion representations for action recognition. The proposed MST-GCN achieves remarkable performance on three challenging benchmark datasets, NTU RGB+D, NTU-120 RGB+D and Kinetics-Skeleton, for skeleton-based action recognition.

Direct-path relative transfer function (DP-RTF) refers to the ratio between the direct-path acoustic transfer functions of two channels. Though DP-RTF fully encodes the sound directional cues and serves as a reliable localization feature, it is often erroneously estimated in the presence of noise and reverberation. This paper proposes a supervised DP-RTF learning method with deep neural networks for robust binaural sound source localization. To exploit the complementarity of single-channel spectrogram and dual-channel difference information, we first recover the direct-path magnitude spectrogram from the contaminated one using a monaural enhancement network, and then predict the DP-RTF from the dual-channel (enhanced-) intensity and phase cues using a binaural enhancement network. In addition, a weighted-matching softmax training loss is designed to promote the predicted DP-RTFs to be concentrated for the same direction and separated for different directions. Finally, the direction of arrival (DOA) of source is estimated by matching the predicted DP-RTF with the ground truths of candidate directions. Experimental results show the superiority of our method for DOA estimation in the environments with various levels of noise and reverberation.

Direct-path relative transfer function (DP-RTF) refers to the ratio between the direct-path acoustic transfer functions of two microphone channels. Though DP-RTF fully encodes the sound spatial cues and serves as a reliable localization feature, it is often erroneously estimated in the presence of noise and reverberation. This paper proposes to learn DP-RTF with deep neural networks for robust binaural sound source localization. A DP-RTF learning network is designed to regress the binaural sensor signals to a real-valued representation of DP-RTF. It consists of a branched convolutional neural network module to separately extract the inter-channel magnitude and phase patterns, and a convolutional recurrent neural network module for joint feature learning. To better explore the speech spectra to aid the DP-RTF estimation, a monaural speech enhancement network is used to recover the direct-path spectrograms from the noisy ones. The enhanced spectrograms are stacked onto the noisy spectrograms to act as the input of the DP-RTF learning network. We train one unique DP-RTF learning network using many different binaural arrays to enable the generalization of DP-RTF learning across arrays. This way avoids time-consuming training data collection and network retraining for a new array, which is very useful in practical application. Experimental results on both simulated and real-world data show the effectiveness of the proposed method for direction of arrival (DOA) estimation in the noisy and reverberant environment, and a good generalization ability to unseen binaural arrays.

The fusion of audio and visual modalities is an important stage of audio-visual speech recognition (AVSR), which is generally approached through feature fusion or decision fusion. Feature fusion can exploit the covariations between features from different modalities effectively, whereas decision fusion shows the robustness of capturing an optimal combination of multimodality. In this work, to take full advantage of the complementarity of the two fusion strategies and address the challenge of inherent ambiguity in noisy environments, we propose a novel hybrid fusion based AVSR method with residual networks and Bidirectional Gated Recurrent Unit (BGRU), which is able to distinguish homophones in both clean and noisy conditions. Specifically, a simple yet effective audio-visual encoder is used to map audio and visual features into a shared latent space to capture more discriminative multi-modal feature and find the internal correlation between spatial-temporal information for different modalities. Furthermore, a decision fusion module is designed to get final predictions in order to robustly utilize the reliability measures of audio-visual information. Finally, we introduce a combined loss, which shows its noise-robustness in learning the joint representation across various modalities. Experimental results on the largest publicly available dataset (LRW) demonstrate the robustness of the proposed method under various noisy conditions.

Cross-modality RGB-infrared (RGB-IR) person reidentification (Re-ID) is a challenging research topic due to the heterogeneity of RGB and infrared images. In this paper, we aim to find some auxiliary modalities, which are homologous with the visible or infrared modalities, to help reduce the modality discrepancy caused by heterogeneous images. Accordingly, a new base-derivative framework is proposed, where base refers to the original visible and infrared modalities, and derivative refers to the two auxiliary modalities that are derived from base. In the proposed framework, the double-modality cross-modal learning problem is reformulated as a four-modality one. After that, the images of all the base and derivative modalities are fed into the feature learning network. With the doubled input images, the learned person features become more discriminative. Furthermore, the proposed framework is optimized by the enhanced intra- and cross-modality constraints with the assistance of two derivative modalities. Experimental results on two publicly available datasets SYSU-MM and RegDB show that the proposed method outperforms the other state-of-the-art methods. For instance, we achieve a gain of over 13 % in terms of both Rank- and mAP on RegDB dataset.

3D speaker tracking using co-located audio-visual sensors has received much attention recently. Though various methods have been attempted to this field, it is still challenging to obtain a reliable 3D tracking result since the position of colocated sensors are restricted to a small area. In this paper, a novel particle filter (PF) based method is proposed for 3D audio-visual speaker tracking. Compared with traditional PF based audio-visual speaker tracking method, our 3D audio-visual tracker has two main characteristics. In the prediction stage, we use audio-visual information at current frame to further adjust the direction of the particles after the particle state transition process, which can make the particles more concentrated around the speaker direction. In the update stage, the particle likelihood is calculated by fusing both the visual distance and audiovisual direction information. Specially, the distance likelihood is obtained according to the camera projection model and the adaptively estimated size of speaker face or head, and the direction likelihood is determined by audio-visual particle fitness. In this way, the particle likelihood can better represent the speaker presence probability in 3D space. Experimental results show that the proposed tracker outperforms other methods and provides a favorable speaker tracking performance both in 3D space and on the image plane.

Asynchronization issue caused by different types of modalities is one of the major problems in audio visual speech recognition (AVSR) research. However, most AVSR systems merely rely on up sampling of video or down sampling of audio to align audio and visual features, assuming that the feature sequences are aligned frame-by-frame. These pre-processing steps oversimplify the asynchrony relation between acoustic signal and lip motion, lacking flexibility and impairing the performance of the system. Although there are systems modeling the asynchrony between the modalities, sometimes they fail to align speech and video precisely over some even all noisy conditions. In this paper, we propose a mutual feature alignment method for AVSR which can make full use of cross modility information to address the asynchronization issue by introducing Mutual Iterative Attention (MIA) mechanism. Our method can automatically learn an alignment in a mutual way by performing mutual attention iteratively between the audio and visual features, relying on the modified encoder structure of Transformer. Experimental results show that our proposed method obtains absolute improvements up to 20.42% over the audio modality alone depending upon the signal-to-noise-ratio (SNR) level. Better recognition performance can also be achieved comparing with the traditional feature concatenation method under both clean and noisy conditions. It is expectable that our proposed mutual feature alignment method can be easily generalized to other multimodal tasks with semantically correlated information.

Lip-reading methods and fusion strategy are crucial for audio-visual speech recognition. In recent years, most approaches involve two separate audio and visual streams with early or late fusion strategies. Such a single-stage fusion method may fail to guarantee the integrity and representativeness of fusion information simultaneously. This paper extends a traditional single-stage fusion network to a two-step feature fusion network by adding an audio-visual early feature fusion (AV-EFF) stream to the baseline model. This method can learn the fusion information of different stages, preserving the original features as much as possible and ensuring the independence of different features. Besides, to capture long-range dependencies of video information, a non-local block is added to the feature extraction part of the visual stream (NL-Visual) to obtain the long-term spatio-temporal features. Experimental results on the two largest public datasets in English (LRW) and Mandarin (LRW-1000) demonstrate our method is superior to other state-of-the-art methods.

Lipreading is an important component of audio-visual speech recognition. However, lips are usually modeled as a whole in lipreading, which ignores that each part of lip focuses on different characteristics of mouth and the overall model can not fit each part perfectly. Besides, features based on the whole lip usually vary a lot according to different speakers, which leads that the training databases usually need to contain as much speakers as possible. In this paper, A part-based lipreading (PBL) method is proposed to deal with the mismatch between an overall lip model and the separate parts of lips, also the excessive dependence of models on the speakers in training set. PBL models lips partly and predicts jointly. It employs a uniform partition strategy on convolutional features and generates several part-level sub-results for final prediction. Experiments are performed on a large publicly available dataset (LRW) and part of it (p-LRW, 65 words), in order to simulate the progressive instructions in the working scene of robots. Word accuracy of PBL reaches 82.8% on LRW and 88.9% on p-LRW. Finally, an end-to-end audio-visual speech recognition system using PBL is established and achieves 98.3% word accuracy on LRW.

Audio-visual speaker tracking in 3D space is a challenging problem. Although the classical particle filter based methods have shown effectiveness in audio-visual speaker tracking, the performance degrades considerably when the measurements are disturbed by noise. To this end, a novel two-layer particle filter is proposed for 3D audio-visual speaker tracking. Firstly, two groups of particles, which are generated from the audio and video streams respectively, are propagated independently in the audio layer and visual layer. Then, the audio and visual likelihoods are combined in an adaptive sigmoid function, which can adjust particle weights according to the confidence of two modalities. Finally, an optimal particle set selected from two groups of particles is proposed to determine the speaker position and reset the particle positions in the next frame. Experiments on AV16.3 database show that our method outperforms the trackers using individual modalities and the existing approaches in the 3D space and on the image plane.

This paper addresses the problem of multiple sound source counting and localization in adverse acoustic environments, using microphone array recordings. The proposed time-frequency (TF) wise spatial spectrum clustering based method contains two stages. First, given the received sensor signals, the spatial correlation matrix is computed and denoised in the TF domain. The TF-wise spatial spectrum is estimated based on the signal subspace information, and further enhanced by an exponential transform, which can increase the reliability of the source presence possibility reflected by spatial spectrum. Second, to jointly count and localize sound sources, the enhanced TF-wise spatial spectra are divided into several clusters with each cluster corresponding to one source. Sources are successively detected by searching the significant peaks of the remaining global spatial spectrum, which is formed using unassigned spatial spectra. After each new source detection, spatial spectra are reassigned to detected sources according to the dominance association between them. The interaction between sources is reduced by iteratively performing new source detection and spatial spectrum assignment. Experiments on both simulated data and real-world data demonstrate the superiority of the proposed method for multiple sound source counting and localization in the environment with different levels of noise and reverberation.

Multiple sound source localization in wireless acoustic sensor networks (WASNs) is a challenging problem. Although compressive sensing based methods have shown effectiveness in uncorrelated sources localization, their performance degrades significantly when they are used to locate multiple speech sources. To this end, we propose a multiple sound source localization method based on the time difference of arrival (TDOA) clustering and the multi-path matching pursuit algorithm. First, TDOAs are calculated locally in time-frequency (TF) bins of sensor recordings. Then, the TDOAs are clustered after utilizing outlier rejection to remove erroneous estimations. Finally, a multi-path matching pursuit algorithm is proposed to solve a sparse localization model for localizing multiple sound sources. Experimental results show that the proposed method yields good performance for multiple sound source localization, especially in strong noisy scenarios.