Method and apparatus for encoding / decoding data representing audio or visual content and related data stream

By using adaptive filtering techniques for decoding and oversampling signals of convolutional matrices, the problems of high computational cost and poor adaptability in existing technologies are solved, and high-quality image reconstruction is achieved.

CN122290616APending Publication Date: 2026-06-26FUNDACIÓN BIKAM
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511945789.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2024-12-24
Filing Date
2025-12-22
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing technologies using artificial neural networks to process oversampled images are computationally expensive and unsuitable for processing images with different downsampling processes, resulting in poor reconstruction quality.

Method used

The first resolution signal is obtained by decoding the first data, the weight is obtained by decoding the second data, and the signal is oversampled by the convolution matrix. The signal quality is improved by combining filtering and adaptive filtering of the convolution matrix to adapt to different downsampling processes.

Benefits of technology

It improves the quality of oversampled signals, bringing them closer to the desired original signal, reduces computational costs, and adapts to image reconstruction using different downsampling processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122290616A_ABST
    Figure CN122290616A_ABST
Patent Text Reader

Abstract

This disclosure relates to methods and apparatuses for encoding / decoding data representing audio or visual content, and related data streams. A method for decoding data representing audio or visual content includes the steps of: - decoding first data (B1) to obtain a signal (IDSdec) at a first resolution; - decoding second data (B2) to obtain multiple weights (Wo); - oversampling the signal at the first resolution (IDSdec) to a signal at a second resolution (IUS) higher than the first resolution; - filtering the signal at the second resolution (IUS) by passing it through at least one convolution of a convolution matrix, wherein at least some coefficients of the convolution matrix are weights in the multiple weights (Wo). A decoding method, apparatus, and associated data stream are also described.
Need to check novelty before this filing date? Find Prior Art