Use of Micro-Modulation Features in Large Vocabulary Continuous Speech Recognition Tasks

Abstract

Most of the state-of-the-art ASR systems take as input a single type of acoustic features, dominated by the traditional feature schemes, i.e., MFCCs or PLPs. However, these features cannot model rapid, intra-frame phenomena present in the actual speech signals. On the other hand, micro-modulation components, inspired by the AM-FM speech model, can capture these important characteristics of spoken speech, resulting in significant performance improvements, as previously shown in small-vocabulary ASR tasks. Yet, they have limited use in large vocabulary ASR applications, where feature post-processing schemes are usually employed. To enable the successful application of these frequency measures in real-life tasks, we investigate their combination with the traditional Cepstral features when employing linear, e.g., HDA, and nonlinear, i.e., bottleneck neural net (BN), feature transforms. This feature combination is investigated in the context of the hybrid DNN-HMM framework, as well. The experimental results reveal that the integration of micro-modulation and Cepstral features, using neural nets, can greatly improve the ASR performance with respect to using the Cepstral features alone. We apply this novel feature extraction approach on different tasks, i.e., a clean speech task (DARPA-WSJ), the Aurora-4 task and a real-life, open-vocabulary, mobile search task, the Speak4it, always reporting improved performance, while the obtained relative word error reduction ranges between 7%-21% depending on the task, e.g., a relative WER improvement of 18% for the Speak4it task, and similar improvements, up to 21%, for the WSJ task are reported.

Creation and Instigation of Triphone based Big-Lexicon Speaker-Independent Continuous Speech Recognition Framework for Kannada Language

“…It was demonstrated that the joined framework, regardless of the low precision of the various levelled TDNN, accomplishes a WRR reduction of 15% according to cutting-edge HMM framework. The author in [7] investigated the exhibition of SR system with the customary Cepstral features when utilizing the linear feature transforms. This combination of features is used to model the DNN-HMM system.…”

Novel Variable Length Energy Separation Algorithm Using Instantaneous Amplitude Features for Replay Detection

“…Here, we have used a linearly-scaled Gabor filterbank to obtain the subband filtered signals. The AM-FM modulation features corresponding to the i th subband filtered signal are extracted from instantaneous frequency fi(t) and amplitude envelope ai(t), where i=1,2,...., L, and L is the number of subband filtered signals [31], i.e.,…”

On the improvement of modulation features using multi-microphone energy tracking for robust distant speech recognition

“…More recently, some of these approaches were revised in the framework of Deep Neural Networks (DNNs) where non-linear modeling is feasible. Networks are trained to extract bottleneck features [5], and combine channels [12], achieving similar or better results compared to beamforming. However, training DNNs on multi-style and multi-channel data [20] is the This research work was supported by the EU under the project I-SUPPORT with grant H2020-643666.…”

“…Their fusion exhibits robustness in noise and mismatch training/testing conditions (e.g., in Aurora-4 task), as indicated by the single-channel ASR results in recent works [5], [16]. However, only a few works [19], [15] examine their performance in reverberant environments.…”