TalonZhang的博客

Title：AN ATTENTION-BASED NEURAL NETWORK APPROACH FOR SINGLE CHANNEL SPEECH ENHANCEMENTWhat’s main claim? Key idea?​ In this paper, the authors explore the attention-based neural network structures for improving the performance of speech enhancement. T

Title：DUAL-PATH RNN: EFFICIENT LONG SEQUENCE MODELING FOR TIME-DOMAIN SINGLE-CHANNEL SPEECH SEPARATIONWhat’s main claim? Key idea?This paper proposes a simple network called dual-path RNN (DPRNN), that organizes any kinds of RNN layers to model long se

Title：T-GSA: TRANSFORMER WITH GAUSSIAN-WEIGHTED SELF-ATTENTION FOR SPEECH ENHANCEMENTWhat’s main claim? Key idea?This paper introduces a self-supervised speech pre-training method called TERA. The authors use a multi-target auxiliary task to pre-train

Title：TERA: Self-Supervised Learning of Transformer Encoder Representation for SpeechWhat’s main claim? Key idea?This paper introduces a self-supervised speech pre-training method called TERA. The authors use a multi-target auxiliary task to pre-train

Paper：Title：Attention Is All You NeedWhat’s main claim? Key idea?​ This paper proposed a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely.Is the idea neat?

Title: SPEAKERFILTER: DEEP LEARNING-BASED TARGET SPEAKER EXTRACTION USING ANCHOR SPEECHWhat’s main claim? Key idea?This paper proposed a novel framework for target speaker separation with anchor speech (Speakerfilter). In the framework, They use a conv