Speech Library

Overview

Speech Library provides a very easy way to work with the end-to-end speech recognition pipeline. The software stack is created to minimize effort required to build speech enabled applications. Speech Library wraps all of the processing blocks and exposes a simple API. The library takes care of proper initialization and data passing between all the components in the pipeline.

Speech Library contains:

• Two core binary libraries, Intel® Feature Extraction library and Intel® Speech Decoder (in lib folder),
• Speech library source code (in the src folder),
• Speech library header files (in the include folder). The library API is in the file speech_library.h. To compile the libraries, please run a .bat/.sh file in the root folder of speech libraries and demos, or run the demonstration script (<INSTALL_DIR>/deployment_tools/demo/speech_recogintion.bat/sh).

Architecture

The implementation of speech recognition pipeline used in demo applications is based on classic HMM/DNN approach.

The pipeline consists of the following stages:

• MFCC feature extraction: the input audio signal or waveform is processed by Intel® Feature Extraction library to create a series of MFCC features.
• Neural acoustic scoring: OpenVINO ™ Inference Engine using neural acoustic model transcribes the extracted features into a sequence of phonemes.
• Language model decoding: the Intel® Speech Decoder takes the phonemes and turns them into text hypothesis. The decoding graph takes into account the grammar of the data, as well as the distribution and probabilities of contiguous specific words (n-grams).

Speech Library API

The Speech Library API was designed to be very simple. It consists of just a few routines which:

• Builds recognizer pipeline
• Provides audio samples for processing
• Inform about new stable recognition result

The flow is described below:

Please refer to: <INSTALL_DIR>/data_processing/audio/speech_recognition/include/speech_library.h to learn more details about the API.

A great example on how to use the API is the source code of offline speech recognition demo.

Running Your Application

Before running compiled binary files, make sure your application can find the Inference Engine, Speech, Decoder and Feature Extraction libraries.

On Linux* operating systems, including Ubuntu*, the LD_LIBRARY_PATH environment variable is usually used to specify directories to be looked for libraries.

You can update the LD_LIBRARY_PATH with paths to the directories in the Inference Engine installation directory where the libraries reside.

Please check run_demo.sh of offline and live speech recognition demos to learn how the LD_LIBRARY_PATH environment parameter can be set.