Deep attractor network for single-microphone speaker separation

Abstract

Despite the overwhelming success of deep learning in various speech processing tasks, the problem of separating simultaneous speakers in a mixture remains challenging. Two major difficulties in such systems are the arbitrary source permutation and unknown number of sources in the mixture. We propose a novel deep learning framework for single channel speech separation by creating attractor points in high dimensional embedding space of the acoustic signals which pull together the time-frequency bins corresponding to each source. Attractor points in this study are created by finding the centroids of the sources in the embedding space, which are subsequently used to determine the similarity of each bin in the mixture to each source. The network is then trained to minimize the reconstruction error of each source by optimizing the embeddings. The proposed model is different from prior works in that it implements an end-to-end training, and it does not depend on the number of sources in the mixture. Two strategies are explored in the test time, K-means and fixed attractor points, where the latter requires no post-processing and can be implemented in real-time. We evaluated our system on Wall Street Journal dataset and show 5.49% improvement over the previous state-of-the-art methods.

Continuous Speech Separation: Dataset and Analysis

“…permutation during model training. Numerous extentions to these methods were proposed with different focuses [3,4,5,6,7,8].…”

Ecological origins of perceptual grouping principles in the auditory system

“…More recent attempts to build computational models of sound segregation similarly focus on the intuitively plausible cue of temporal coincidence [24,25]. Current state-of-the-art engineering methods instead rely on learning how to group acoustic energy from labeled sound mixtures [44,45], but are at present difficult to probe for insight into the underlying acoustic dependencies. Our methodology falls between these two traditions, utilizing the rich set of constraints imposed by natural signals but providing interpretable insight into factors that might underlie grouping.…”

Practical Applicability of Deep Neural Networks for Overlapping Speaker Separation

“…These vectors are then used to generate masks to filter out the individual speakers from the mixture. By using these intermediate embedding vectors instead of directly outputting the masks, the so called permutation problem [11] is avoided. This paper is organised as follows.…”