Search

Deep neural network (DNN) based end-to-end optimization in the complex time-frequency (T-F) domain or time domain has shown considerable potential in monaural speech separation. Many recent studies optimize loss functions defined solely in the time or complex domain, without including a loss on magnitude. Although such loss functions typically produce better scores if the evaluation metrics are objective time-domain metrics, they however produce worse scores on speech quality and intelligibility metrics and usually lead to worse speech recognition performance, compared with including a loss on magnitude. While this phenomenon has been experimentally observed by many studies, it is often not accurately explained and there lacks a thorough understanding on its fundamental cause. This letter provides a novel view from the perspective of the implicit compensation between estimated magnitude and phase. Analytical results based on monaural speech separation and robust automatic speech recognition (ASR) tasks in noisy-reverberant conditions support the validity of our view. [ABSTRACT FROM AUTHOR]

Speech enhancement and source separation systems based on deep learning have recently reached unprecedented levels of quality, to the point that performance is reaching a new ceiling. Most systems rely on estimating the magnitude of a target source by estimating a real-valued mask to be applied to a time-frequency representation of the mixture signal. A limiting factor in such approaches is a lack of phase estimation: the phase of the mixture is most often used when reconstructing the estimated time-domain signal. Here, we propose “magbook,” “phasebook,” and “combook,” three new types of layers based on discrete representations that can be used to estimate complex time-frequency masks. Magbook layers extend classical sigmoidal units and a recently introduced convex softmax activation for mask-based magnitude estimation. Phasebook layers use a similar structure to give an estimate of the phase mask without suffering from phase wrapping issues. Combook layers are an alternative to the magbook–phasebook combination that directly estimate complex masks. We present various training and inference schemes involving these representations, and explain in particular how to include them in an end-to-end learning framework. We also present an oracle study to assess upper bounds on performance for various types of masks using discrete phase representations. We evaluate the proposed methods on the wsj0-2mix dataset, a well-studied corpus for single-channel speaker-independent speaker separation, matching the performance of state-of-the-art mask-based approaches without requiring additional phase reconstruction steps. [ABSTRACT FROM AUTHOR]

We propose a method for characterizing sound activity in fixed spaces through segmentation, indexing, and retrieval of continuous audio recordings. Regarding segmentation, we present a dynamic Bayesian network (DBN) that jointly infers onsets and end times of the most prominent sound events in the space, along with an extension of the algorithm for covering large spaces with distributed microphone arrays. Each segmented sound event is indexed with a hidden Markov model (HMM) that models the distribution of example-based queries that a user would employ to retrieve the event (or similar events). In order to increase the efficiency of the retrieval search, we recursively apply a modified spectral clustering algorithm to group similar sound events based on the distance between their corresponding HMMs. We then conduct a formal user study to obtain the relevancy decisions necessary for evaluation of our retrieval algorithm on both automatically and manually segmented sound clips. Furthermore, our segmentation and retrieval algorithms are shown to be effective in both quiet indoor and noisy outdoor recording conditions. [ABSTRACT FROM AUTHOR]

Organizing a database of user-contributed environmental sound recordings allows sound files to be linked not only by the semantic tags and labels applied to them, but also to other sounds with similar acoustic characteristics. Of paramount importance in navigating these databases are the problems of retrieving similar sounds using text- or sound-based queries, and automatically annotating unlabeled sounds. We propose an integrated system, which can be used for text-based retrieval of unlabeled audio, content-based query-by-example, and automatic annotation of unlabeled sound files. To this end, we introduce an ontological framework where sounds are connected to each other based on the similarity between acoustic features specifically adapted to environmental sounds, while semantic tags and sounds are connected through link weights that are optimized based on user provided tags. Furthermore, tags are linked to each other through a measure of semantic similarity, which allows for efficient incorporation of out-of-vocabulary tags, that is, tags that do not yet exist in the database. Results on two freely available databases of environmental sounds contributed and labeled by non expert users demonstrate effective recall, precision, and average precision scores for both the text-based retrieval and annotation tasks. [ABSTRACT FROM AUTHOR]