Source Coding With Time Advanced Excitation Signal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contemporary speech coding methods face challenges in accurately modeling the excitation signal, leading to potential peaks in prediction error and inefficiencies in bit-rate management, particularly when dealing with variable speech segments and the need for flexible pulse allocation within frames.
Innovation Solution
The introduction of time advanced excitation generation, where the excitation signal is determined to partly cover the next frame or sub-frame, allowing for improved modeling by doubling the number of pulses in a frame without increasing coding delay, and utilizing this approach in both fixed-rate and variable-rate coders.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the excitation signal is determined only for the current frame, then the coding delay is kept low, but the modeling accuracy of the excitation signal deteriorates
Solution Approach 1:
The patent applies preliminary action by determining the excitation signal to cover not only the current frame but also extending into the next frame or sub-frame. This advance determination improves modeling accuracy by capturing temporal dependencies that span frame boundaries, while the look-ahead buffer manages the timing aspects to minimize coding delay impact.
2Measurement precision
If the number of pulses in a frame is increased to improve speech representation, then the speech quality improves, but the bit rate increases
Solution Approach 1:
The patent applies local quality by allowing variable pulse allocation within frames based on local speech characteristics. The look-ahead mechanism identifies regions requiring higher representation accuracy (such as areas with LP residual peaks) and allocates pulses selectively to those regions, rather than uniformly increasing pulse count across all frames, thus maintaining speech quality while controlling overall bit rate.
3Device complexity
If the excitation signal modeling is simplified to reduce system complexity, then the system complexity decreases, but the error signal modeling accuracy deteriorates
Solution Approach 1:
The patent uses preliminary action through the look-ahead buffer to pre-determine excitation signals that extend into future frames. This approach improves error signal modeling accuracy by capturing temporal dependencies without requiring complex real-time processing during the coding phase, thus achieving better accuracy while maintaining relatively simple system complexity.
Data Source
AI summary
A method and a device for source coding with a time advanced excitation signal. During an encoding process, a source data signal is first divided into consecutive blocks, then a first set of parameters related to a filter describing properties of a first block covering a first time period is extracted, followed by the extraction of a second set of parameters related to an excitation signal for said filter, where said second set of parameters is determined from and describing properties of both the first block and a second block following the first block within a second time period starting later than said first time period and extending outside said first time period.