Open Source Speech Software

a tool for segmenting, labeling and transcribing speech

MARF is a general cross-platform framework with a collection of algorithms for audio (voice, speech, and sound) and natural language text analysis and recognition along with sample applications (identification, NLP, etc.) of its use, implemented in Java.

XVoice provides voice control of X applications using IBM's ViaVoice for Linux (free download at their web site). Both user-defined commands and dictation are supported. It can be used to write letters, write code, control netscape, etc.

Advanced Speech Signal Analysis library provides a structure to handle various file formats and a variety of analysis functions commonly used in speech processing.

The Blind Audio Tactile Mapping System (BATS) attempts to address the lack of spatial information available for visually impaired students.

ASR-Builder provides an easy-to-use interface to the HTK toolkit, that allows users to build ASR systems. ASR-Builder provides a platform that performs house-keeping tasks when using HTK and also provides default training/testing/recognition scripts.

Cotovía is a unit-selection text-to-speech system for Galician and Spanish. Cotovía is distributed under the GPL3.0+ license, but each of the avaliable speaker voices has its own license. The speakers available at sourceforge are free for commercial and non-commercial uses. Another speaker, free for non-commercial uses, is avaliable through external links (see the Blog section). Cotovia has been developed by the University de Vigo and the center 'Ramón Piñeiro' for Research in Humanities, both in Galicia, Spain. Its development has involved a research group of linguists and engineers. Cotovía has been developed as a research project, therefore most of the work has been focused on the most interesting aspects from a scientific point of view. Although the performance of the whole TTS system is quite good, there are some parts that could be clearly improved. Cotovia files and installing instructions are available at the Files and Git sections.

EBBA is a project aiming to develop an advanced chatbot by combining AIML, 3d facial expressions, speech synthesizer, speech recognition and an iq-test solving functionality.

ILA stands for (kind of) intelligent, learning assistant and is a speech recognition system aka voice assistant very similar to Siri, Google Now and Cortana. ILA is fully customizable and you can teach her/him/it new things by yourself like executing system commands, opening web pages, programs and apps or just some basic conversation :-) ILA runs on Java und thus is compatible to Windows, Mac and Linux. It is designed to integrate with your home enviroment and for example build up your own, free and open Amazon Echo replacement ;-) Right now the key components of ILA are the open source speech recognition CMU Sphinx-4, Google (Speech Recognition/Text-To-Speech) and MaryTTS (Text-To-Speech). The goal is to make ILA completely free of Google by improving all aspects of the open source systems. Since version 3.3 users can also write own add-ons to extend ILA. ILA's successor is the SEPIA Framework: https://sepia-framework.github.io/ Hope you enjoy ILA - Florian

This project will show how to implement the Hidden Markov Model approximations of Voice Recognition into embedded and low power systems.

Find the pitch of a power spectrum (signal) as per the afferent/efferent neural crossover. This occurs between the Lateral olivocochlear efferents and the inner hair cell afferents.

Matsig is an object-oriented signal class library (Toolbox in MATLAB lingo) for MATLAB 6.5 and later. It implements a signal class, simplifying operations and manipulations common in audio signal processing and speech processing.

An open PHP-based framework for the Nabaztag™ (http://www.nabaztag.com/) electronic pet. Due to major changes on the Violet backend, OpenNab can no longer be connected to it but it still can be used as a standalone server to set your bunnies free !

Platform for Annotated Corpora in XML Integrated tool for corpus linguists built on Eclipse, Vex, Subversive, etc. for creating and editing transcriptions and annotations, querying, managing version controlled data, and building a shippable corpus.

SAMI is an extensible, voice-controlled home automation system which seamlessly controls all of the pieces of your apartment or house, without installation or monthly fees. For more information about how to use her, see the documentation tab!

VoxForge collects user-submitted speech audio files for the creation of Acoustic Models for Free and Open Source Speech Recognition Engines such as HTK, Julius, ISIP and Sphinx.

Software to fit whole-sentence language models using the principle of maximum entropy. For developers of speech recognizers, text prediction interfaces, OCR, machine translation software.

jATLAS is a Java implementation of ATLAS [Architecture and Tools for Linguistic Analysis Systems]. For more information, see http://jatlas.sourceforge.net.

Spantus - signal(speech) analytical tool

Yet Another Audio Feature Extractor is a toolbox for audio analysis. Easy to use and efficient at extracting a large number of audio features simultaneously. WAV and MP3 files supported, or embedding in C++, Python or Matlab applications.