17836390. Automatic Speech Recognition Systems and Processes simplified abstract (Microsoft Technology Licensing, LLC)

Automatic Speech Recognition Systems and Processes

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

Kshitiz Kumar of Redmond WA (US)

Jian Wu of Bellevue WA (US)

Bo Ren of Bellevue WA (US)

Tianyu Wu of Suzhou (CN)

Fahimeh Bahmaninezhad of San Mateo CA (US)

Edward C. Lin of Beijing (CN)

Xiaoyang Chen of Jiangsu (CN)

Changliang Liu of Bellevue WA (US)

Automatic Speech Recognition Systems and Processes - A simplified explanation of the abstract

This abstract first appeared for US patent application 17836390 titled 'Automatic Speech Recognition Systems and Processes

Simplified Explanation

Abstract: A data processing system is designed to receive speech data in multiple languages and convert it into letters. The system then applies linguistic rules to normalize the speech data for Latin-based languages, builds a computer model using the normalized data, fine-tunes the model with additional speech data, and finally recognizes words in a target language using the refined model.

Patent/Innovation Explanation:

• Data processing system for speech data in multiple languages
• Converts speech data into letters
• Applies linguistic rules to normalize speech data for Latin-based languages
• Builds a computer model using normalized speech data
• Fine-tunes the computer model with additional speech data
• Recognizes words in a target language using the refined computer model

Potential Applications:

• Speech recognition systems for multilingual environments
• Language learning applications
• Translation services
• Voice-controlled devices and virtual assistants

Problems Solved:

• Efficiently processing speech data in multiple languages
• Normalizing speech data for Latin-based languages using linguistic rules
• Building accurate computer models for speech recognition
• Improving the recognition of words in a target language

Benefits:

• Enhanced accuracy and efficiency in speech recognition
• Improved language learning experiences
• Seamless translation services
• More reliable voice-controlled devices and virtual assistants

Original Abstract Submitted

A data processing system is implemented for receiving speech data for a plurality of languages, and determining letters from the speech data. The data processing system also implements normalizing the speech data by applying linguistic based rules for Latin-based languages on the determined letters, building a computer model using the normalized speech data, fine-tuning the computer model using additional speech data, and recognizing words in a target language using the fine-tuned computer model.