Hemera - Intelligent System

About

Hemera is a Virtual Intelligent System aggregating some more advanced Artificial Intelligence Technologies (speech, speech recognition, form recognition ...).
Final aim is applications in day tasks, domotics and robotics ...

Modus Operandi

Nowadays, Artificial Intelligence Technologies have impressively evolved (and they will continue !).
There is lots of system based on such technologies, but generally specialized in one domain.
The challenge of Hemera project is to aggregate enough Artificial Intelligence Technologies producing an intelligent system with abilities like speech, speech recognition, form recognition, virtual representation, learning, problem solving ...

Policy

Project policy is to focus on open-source projects to benefit from the efforts of communities and participate in making feedback.
Nevertheless, in rare cases, proprietary third-party tools may be used (will be mentioned).

This is an international project, everything is done to allow to easily take care of new languages.
Initially, the targets are English and French.

License

Hemera is developed under the GNU General Public License version 3.0 (GPLv3).

Inspiration

In Greek mythology Hemera was the personification of day and one of the Protogenoi or primordial deities.

Quick Start

See Hemera Quick Start.

Contact

You can contact us on Hemera Forum.
You can send me an e-mail.

Guides

Documentation

Installation & Upgrade

Help and Tutorials

Development and Build Tools

This is the list of used development and build tools.
A great thanks for corresponding projects' members for their work.

Command language interpreter: GNU/Bash
Version control system: Apache Subversion
Object programming language: Java openJDK version or Oracle/Sun version
Build tool (Java-based): Apache Ant
Integrated Development Environment (Java-based): Eclipse
Tomcat "Launcher": Sysdeo Eclipse plugin

Third-party Tools

This is the list of used third-party tools, and how they are used.
A great thanks for corresponding projects' members for their work.

General

Hemera project logo comes from Wikimedia Commons, source

FileSystem events monitor: inotify-tools
Speech sound player/recorder/converter: SoX
Speech sound player (alternate): aplay
Various tools: GNU Which

Web Services Module

Java Servlet container: Apache Tomcat
Java Servlet framework: Apache Struts
Java Command-pattern framework: XWork
Java library for transforming various elements to JSON: JSON-lib
Template engine: FreeMarker
Java logging services: Apache Log4j
Java Bytecode manipulator: Javassist
Object-Graph Navigation Language: OGNL
Fast and concise JavaScript Library: jQuery
Charting library written in pure JavaScript: Highcharts

Speech core Module

Speech synthesizer/Phoneme mnemonics writer: eSpeak
(proprietary) Speech synthesizer: MBROLA Project

Speech Recognition core Module

Speech recognition processor

Speech Recognition Toolkit: CMU Sphinx
Speech Recognition models (French): LIUM Tools

Speech recognition models creation

Language model toolkit: SRILM Tools
Text phonetization (French): LIA_PHON Tool

Special Thanks

Special thanks to:

Yannick Estève for crucial information about Speech Recognition (mainly sphinx3 configuration for use with LIUM models and lexical/language models creation)

Main Page

Contents