ScanSoft offers free evaluation of speech SDK for games

Speaker-independent voice control or interaction with characters

In June, ScanSoft announced that its family of VoCon Gaming Speech Software Developer Kit (SDK) products is now available for free download and evaluation for qualified PC, PlayStation 2, and GameCube developers. Games such as SOCOM: U.S. Navy SEALs for Sony PlayStation 2 and Mario Party 6 for Nintendo GameCube have used speech recognition technology from the VoCon Games Speech SDK. The speaker-independent software uses phoneme-based speech recognition. Developers specify the allowed commands at any point in the game with a structured grammar.

The speech recognition software can provide command and control of the gaming environment in addition to interactive speech in the game. While speech is not well-suited for quick reaction commands, it can supplement those commands with environmental control, such as changing weapons without memorizing complex control sequences. It is also natural when speech is the way that the characters would communicate if the situation were real, such as commanders issuing orders or simulated characters interacting in a social situation.

The game market is crowded and very competitive, and speech control gives the games a new dimension. There is also the potential for a game to reach a broader (and perhaps older) audience if interfaces were more intuitive and didn’t require mastering complex manual control actions. The free evaluation kit will give game developers a way of testing voice interfaces without committing to a purchase.

Alan Schwartz, vice president, Embedded Speech, SpeechWorks Solutions from ScanSoft, said, “By making our Speech Games SDK available for developers, we’re providing developers with the opportunity to explore the power of speech, and expand its opportunity for gaming titles on current and future generations of PC, handheld, and console platforms.”

Acoustic models for major European and Asian languages are available in addition to English. The SDK also includes models for children’s voices as well as male and female adult voices.