Sensory, Inc., the world leader in embedded speech technologies, released the latest member of its award-winning speech recognition integrated circuit (IC) family, RSC-4192O, with one time programmable (OTP) memory aimed at consumer product markets where low cost and high quality are essential. To compliment the new IC, Sensory also released an enhanced version of its firmware, FluentChip(TM) 3.0, to improve noise-immunity, speech and music synthesis, and multi-language speech recognition accuracy.
Industry analyst Bill Meisel of TMA Associates states, “As consumer electronic devices get increasingly small and more complex, voice user interface (VUI) – with speech recognition at the core – will play a key and necessary role in creating a positive and more intuitive user experience. Sensory has maintained majority market share in the IC-based speech recognition market for many years, and the RSC-4192O with FluentChip 3.0 helps to further strengthen its lead by helping manufacturers incorporate better, more cost-effective speech recognition applications into their products.”
“Sensory is dedicated to providing the highest accuracy speech recognition solutions at the lowest system cost. The programmability of the RSC-4192O speeds time-to-market and reduces minimum order quantities, opening up new markets for speech recognition control,” said Todd Mozer, CEO of Sensory, Inc. “On top of that, FluentChip(TM) 3.0 is designed to answer the real-world demand for noise-robustness in speech recognition applications for consumer devices such as phones, wireless devices and interactive toys.”
RSC-4192O With OTP Memory
RSC-4192O joins the RSC-4128 and RSC-464, the most widely deployed speech recognition IC family in history, integrating speech-optimized digital and analog processing blocks into a single chip solution capable of accurate speech recognition as well as high-quality, low data-rate compressed speech. Based on an 8-bit general purpose microcontroller, the RSC-4192O offers 192Kbytes of on-chip OTP memory with cost-effective volume pricing and lightening-fast lead times. The RSC-4192O is a completely self-contained speech input/output (I/O) system with on-chip ADC, DAC, output amplification, and everything expected from a state-of-the-art 8-bit microcontroller.
FluentChip 3.0 for Noisy, Real-World Environments
The RSC-4x IC family supports FluentChip firmware which includes advanced algorithms that add features and increase accuracy. Sensory has added new capabilities to FluentChip(TM) 3.0 including:
- Enhanced speech features which dramatically improve recognition accuracy in high-noise environments typical in homes, automobiles and industrial spaces
- New international lexicons and pronunciation predictors to improve accuracy over the many thousands of words typical in the world’s diverse languages
- Real-time LipSync, enabling robotic animation such that the character’s mouth will move synchronously and accurately in real time with the user’s speech
- New acoustic models with 50 percent smaller code size, freeing much more of the RSC-4x processors’ code space for technology features in the end product
FluentChip 3.0 includes Sensory’s unique T2SI(TM) (text-to-speaker-independent) technology enabling manufacturers to program command sets in minutes vs. months with no custom training required. FluentChip 3.0 also includes SD (speaker dependent; trained to one voice), T2SISD (allows both speaker independent and speaker dependent customized commands in the same command set), and SV (speaker verification; voice biometric password) speech recognition. A whole suite of technologies for enabling electro-mechanical animation is also included, such as Real-time LipSync, SoundSource tracking, and Beat Predict for dancing. FluentChip 3.0 also offers speech synthesis and MIDI-like music synthesis capabilities. Specialized tools are offered to facilitate fast prototyping and product development for members of the RSC-4x family, including the RSC-4192O.
New Development Tools Available
Sensory offers a low-cost toolkit in order to preview and develop its speech technologies in a real-world environment. The new RSC-4x Demo/Evaluation Toolkit V2 includes an RSC-4x-based evaluation board with upgrades such as a USB interface, 0 wait states for improved quality synthesis, and 32MBytes of serial flash memory for storing synthesized speech. It comes bundled with FluentChip 3.0 technology and is ready to support the RSC-4192O. The new toolkit is available to order from the Sensory website.
A new programmer tool from Phyton enables programming the RSC-4192O in 100LQFP package form at the customer’s lab. This complements Phyton’s existing suite of integrated development environment (IDE), C compiler and debugger, and emulator for the RSC-4x family.
For developers who wish to incorporate RSC-4x capabilities into smaller scale projects, Sensory’s new 40-pin DIP footprint VR Stamp(TM) module also supports the FluentChip 3.0 technology library. Sensory’s VR Stamp reduces development effort by incorporating most system design features, promoting rapid deployment of speech technologies into consumer electronic products. The VR Stamp Toolkit provides a development environment for creating, programming and experimenting with code. More information is available on the Sensory website.
About Sensory, Inc.
Sensory, Inc. is the world leader in embedded speech technologies. Sensory is a profitable private company offering a complete line of IC and software-only solutions for speech recognition, speech synthesis, speaker verification, music synthesis and more. The company’s products are widely deployed in consumer electronics applications including telephones, home automation, toys, remote controls, automotive, security, and learning aids. Sensory’s customers represent the leaders in consumer electronics, including such companies as Hasbro, JVC, Kenwood, Mattel, Mitsubishi, Toshiba, Uniden and Sony. Headquartered in Sunnyvale, CA, additional Sensory offices are located in Portland, Hong Kong, Tokyo and Vienna.