SpeechIO
- Creator: L2F/INESC-ID email
- Publisher: Rage project
- Owner: L2F/INESC-ID email
Multimodal Natural Language Processing, suitable for games.
Short non-technical description: SpeechIO provides tools for game developers to process multimodal Natural Language inputs, in support of some form of understanding of a gamer's intention, as expressed in its own words, spoken or written.
Technical description:
SpeechIO computes various interpretations of Natural Language. It accepts plain text or raw audio, applies a series of transformations to it, and outputs a different view of the input, also text or speech, by classification on a predefined set of categories or by conversion to other Natural Language representation. It is intended to transcribe speech, sentiment analysis, question answering, or a combination of these that pursue some form of automated Natural Language Understanding. This component describes frontend/client access to external services where its core functionalities are deployed.
The component is available "as is" without warranties or conditions of any kind. The authors of the software are not planning to provide any kind of technical support to third parties on any issue they may find with the component.
Language: English
Access URL: https://bitbucket.org/L2F-INESCID/speechio
Download: SpeechIO.zip
speech processing
natural language understanding
domain adaptation
Source code:
Documentation:
Setup files:
Version: 0.7
Version notes: Initial version of a RAGE compliant code base. External services are all fully functional, but not all are yet exposed in the asset.
Development status: Under Development
Commit URL: https://bitbucket.org/L2F-INESCID/speechio/commits/all
Type: Apache 2.0 (Apache License 2.0)
URL: https://bitbucket.org/L2F-INESCID/speechio/raw/642612fa472a28c8e276495fe851eacb7c20a01b/License.txt
