Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks

Abstract

Convolutional Neural Networks (CNNs) are effective models for reducing spectral variations and modeling spectral correlations in acoustic features for automatic speech recognition (ASR). Hybrid speech recognition systems incorporating CNNs with Hidden Markov Models/Gaussian Mixture Models (HMMs/GMMs) have achieved the state-of-the-art in various benchmarks. Meanwhile, Connectionist Temporal Classification (CTC) with Recurrent Neural Networks (RNNs), which is proposed for labeling unsegmented sequences, makes it feasible to train an end-to-end speech recognition system instead of hybrid settings. However, RNNs are computationally expensive and sometimes difficult to train. In this paper, inspired by the advantages of both CNNs and the CTC approach, we propose an end-to-end speech framework for sequence labeling, by combining hierarchical CNNs with CTC directly without recurrent connections. By evaluating the approach on the TIMIT phoneme recognition task, we show that the proposed model is not only computationally efficient, but also competitive with the existing baseline systems. Moreover, we argue that CNNs have the capability to model temporal correlations with appropriate context information.

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• Sequence Transduction with Recurrent Neural Networks
• Theano: new features and speed improvements
• cuDNN: Efficient Primitives for Deep Learning
• A Simple Way to Initialize Recurrent Networks of Rectified Linear Units
• EESEN: End-to-End Speech Recognition using Deep RNN Models and WFST-based Decoding
• Blocks and Fuel: Frameworks for deep learning

Cited by in corpus (16)

• The History Began from AlexNet: A Comprehensive Survey on Deep Learning Approaches
• XNOR Neural Engine: a Hardware Accelerator IP for 21.6 fJ/op Binary Neural Network Inference
• Efficient Processing of Deep Neural Networks: A Tutorial and Survey
• Sequence Modeling via Segmentations
• Who Needs Words? Lexicon-Free Speech Recognition
• Advances in Joint CTC-Attention based End-to-End Speech Recognition with a Deep CNN Encoder and RNN-LM
• Speech recognition with quaternion neural networks
• Attention Control with Metric Learning Alignment for Image Set-based Recognition
• Dependency-aware Attention Control for Unconstrained Face Recognition with Image Sets
• Effects of Number of Filters of Convolutional Layers on Speech Recognition Model Accuracy
• End to End ASR System with Automatic Punctuation Insertion
• A Novel Approach for Effective Learning in Low Resourced Scenarios
• Novel Perception Algorithmic Framework For Object Identification and Tracking In Autonomous Navigation
• Robustness of end-to-end Automatic Speech Recognition Models -- A Case Study using Mozilla DeepSpeech
• Time-Frequency Localization Using Deep Convolutional Maxout Neural Network in Persian Speech Recognition
• WaDeNet: Wavelet Decomposition based CNN for Speech Processing