Perceptral recently filed a patent application for its proprietary method of Acoustic Perceptual Event (APE) segmentation. The patent covers its “Acoustic Perceptual Analysis and Synthesis System.”

APE segmentation defines and manipulates units of sound, including the handling of speech prosody (accent, intonation, timing, and  timbre)  at a finer level. APE segmentation is based on identifying perceptually salient features that underlie transient sound events—classifying, measuring, and comparing them, to create an entirely novel foundation for sound recognition, analysis, and synthesis. Currently established technologies approach sound essentially as a sequence of relatively disjunct, equivalently-valued steady states; Perceptral’s methods re-characterize sound as a network of interrelated dynamic events with weighted values determined by the contribution of each to a meaningful percept by the listener.

Perceptral is an early-stage Research & Development firm, established in May 2008 in Racine, Wis. To date, the company has been selected for four Small Business Innovation Research (SBIR) awards from the Department of Defense. Together these contracts represent nearly $970,000.

The mission of Perceptral is to revolutionize the handling of speech prosody in digital sound applications. Current and planned efforts address:

• Variable speed speech synthesis;
• Animated voices and sound blends for entertainment and presentations;
• Sound compression and decoding;
• Forensic reconstruction of degraded audio;
• Dialect/language identification and classification;
• Pedagogical tools for accent reduction and acquisition;
• Automated dubbing in second accent/dialect/language; and
• Voice verification/identification/security. 

For more information, contact Perceptral at (262) 898-1258, email info@perceptral.com, or visit the company Web site at www.perceptral.com.