Tencent Technology (Shenzhen) Company Limited (20240212671). SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM simplified abstract
Contents
- 1 SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Key Features and Innovation
- 1.6 Potential Applications
- 1.7 Problems Solved
- 1.8 Benefits
- 1.9 Commercial Applications
- 1.10 Prior Art
- 1.11 Frequently Updated Research
- 1.12 Questions about Speech Decoding Technology
- 1.13 Original Abstract Submitted
SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM
Organization Name
Tencent Technology (Shenzhen) Company Limited
Inventor(s)
Xiaozheng Jian of Shenzhen (CN)
SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240212671 titled 'SPEECH DECODING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM
Simplified Explanation
The patent application describes a method for speech decoding using computer devices. It involves decoding audio data, determining pruning parameters, and decoding subsequent audio frames.
- Obtaining audio data corresponding to speech, including multiple audio frames.
- Decoding the first audio frame using two decoding networks to obtain tokens with decoding scores.
- Determining pruning parameters based on the token with the smallest decoding score.
- Using the pruning parameters to restrict the decoding process of the second audio frame.
- Decoding the second audio frame using the decoding networks and the pruning parameters.
Key Features and Innovation
- Utilizes multiple decoding networks for speech decoding. - Implements pruning parameters to optimize the decoding process. - Enhances accuracy and efficiency in speech recognition.
Potential Applications
This technology can be applied in: - Speech recognition software. - Language translation tools. - Voice-controlled devices.
Problems Solved
- Improves the accuracy of speech decoding. - Enhances the efficiency of language processing. - Optimizes the performance of speech recognition systems.
Benefits
- Increased accuracy in speech recognition. - Faster language processing. - Enhanced user experience with voice-controlled devices.
Commercial Applications
Title: Advanced Speech Decoding Technology for Enhanced Language Processing This technology can be utilized in: - Virtual assistants. - Call center automation. - Language learning applications.
Prior Art
Further research can be conducted in the field of speech decoding and language processing to explore similar technologies and advancements.
Frequently Updated Research
Stay updated on advancements in speech recognition technology and language processing to enhance the efficiency and accuracy of speech decoding systems.
Questions about Speech Decoding Technology
1. How does this technology improve the accuracy of speech recognition?
- This technology enhances accuracy by utilizing multiple decoding networks and pruning parameters to optimize the decoding process.
2. What are the potential commercial applications of this speech decoding technology?
- This technology can be applied in virtual assistants, call center automation, and language learning applications.
Original Abstract Submitted
a method for speech decoding is performed by a computer device. the method includes: obtaining audio data corresponding to a speech, the audio data including a first audio frame and a second audio frame; decoding the first audio frame using a first decoding network corresponding to a low-order language model and a second decoding network corresponding to a differential language model to obtain a plurality of first tokens, each first token having a corresponding decoding score according to the first and second decoding network; determining pruning parameters according to a target token of the plurality of first tokens having a smallest decoding score, wherein the pruning parameters is used for restricting a decoding process of the second audio frame; and decoding the second audio frame using the first decoding network and the second decoding network according to the first token list and the pruning parameters.
