Speech Encoding Using Pre-Encoded Database
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech coding techniques, such as Adaptive Multi-Rate Wideband (AMR-WB) and Enhanced Voice Services (EVS), consume high resources like processor and memory due to their complex operations, which can impact performance and user experience.
Innovation Solution
Implementing a method that uses a pre-encoded database with a machine learning scheme, like a neural network classifier, to generate a bitstream compatible with speech coding standards without actual run-time encoding, reducing complexity and resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional speech coding techniques (AMR-WB, EVS) are used for real-time encoding, then coding standard compatibility is achieved, but processor usage and memory usage increase significantly
Solution Approach 1:
The patent pre-encodes speech data offline using standard coding techniques (AMR-WB, EVS) to create a compressed database. This preliminary encoding action transfers the computationally intensive work from runtime to offline processing, allowing runtime operations to only perform database searching and parameter extraction, thereby maintaining coding standard compatibility while dramatically reducing runtime processor and memory usage.
2Manufacturing precision
If traditional run-time encoding is performed, then accurate speech compression is achieved, but processing time increases
Solution Approach 1:
The patent performs speech encoding in advance during an offline phase, creating a pre-encoded database with compressed speech data. At runtime, the system only needs to search this pre-encoded database and extract relevant parameters, which dramatically reduces processing time while maintaining speech compression accuracy through the use of machine learning models trained on the pre-encoded data.
Data Source
AI summary
Methods, systems, and devices for encoding are described. A device, which may be otherwise known as user equipment (UE), may support standards-compatible audio encoding (e.g., speech encoding) using a pre-encoded database. The device may receive a digital representation of an audio signal and identify, based on receiving the digital representation of the audio signal, a database that is pre-encoded according to a coding standard and that includes a quantity of digital representations of other audio signals. The device may encode the digital representation of the audio signal using a machine learning scheme and information from the database pre-encoded according to the coding standard. The device may generate a bitstream of the digital representation that is compatible with the coding standard based on encoding the digital representation of the audio signal, and output a representation of the bitstream.