Obtaining a target representation of a time sample of a signal

Patent metadata

Field	Value
Publication number	US-11265553-B2
Application number	US-201816648107-A
Country	US
Kind code	B2
Filing date	Sep 21, 2018
Priority date	Sep 22, 2017
Publication date	Mar 1, 2022
Grant date	Mar 1, 2022

Abstract

A first value of a first data element ( 311 ) in a first set of data elements ( 310 ) is obtained, the first set of data elements being based on a first time sample of a signal. A second value of a second data element ( 321 ) in a second set of data elements ( 320 ) is obtained, the second set of data elements being based on a second, later time sample of the signal. A measure of similarity is derived between the first value and the second value. Based on the derived measure, a quantisation parameter useable in performing quantisation on data based on the first time sample of the signal is determined. Output data is generated using the quantisation parameter.

First claim

The invention claimed is:

1. A method for obtaining a target representation of a first time sample of a signal, the signal being arranged in accordance with a tiered hierarchy having multiple levels of quality, the method comprising:
comparing a first entropy associated with first reconstruction data useable by a decoder device to obtain the target representation using a first processing mode with a second entropy associated with second reconstruction data useable by the decoder device to obtain the target representation using a second, different processing mode, the target representation being at a first level of quality in the tiered hierarchy;
determining whether the target representation is to be obtained by the decoder device using the first processing mode or the second processing mode based on a result of the comparison, wherein the first processing mode is to be used when the result of the comparison indicates that the first entropy is lower than the second entropy and the second processing mode is to be used when the result of the comparison indicates that the second entropy is lower than the first entropy; and
outputting data arranged to indicate whether the first processing mode or the second processing mode is to be used by the decoder device to obtain the target representation,
wherein using the first processing mode to obtain the target representation comprises the first reconstruction data and a further representation of the first time sample of the signal that is at a second, lower level of quality in the tiered hierarchy being processed by the decoder device, and
wherein using the second processing mode to obtain the target representation comprises the second reconstruction data and a representation of a second, different time sample of the signal being processed by the decoder device.

2. A method according to claim 1 , wherein the target representation and the representation of the second time sample are at a same level of quality as each other in the tiered hierarchy.

3. A method according to claim 1 , the method comprising receiving data arranged to indicate one or more of the first entropy and the second entropy.

4. A method according to claim 1 , the method comprising:
determining the first entropy by deriving the first reconstruction data, including: performing an upsampling operation on the further representation to derive a preliminary representation of the first time sample of the signal at the first level of quality; and
deriving the first reconstruction data using the preliminary representation.

5. A method according to claim 1 , wherein one or more of the first reconstruction data and the second reconstruction data comprises intra-residual data useable by the decoder device to obtain the target representation.

6. A method according to claim 5 , wherein the inter-processing data for the second reconstruction data comprises motion compensation data relating to motion of one or more items between the representation of the second time sample and the target representation.

7. A method according to claim 6 , wherein the inter-processing data for the second reconstruction data comprises inter-residual data derived based on a comparison between a predicted representation of the first time sample of the signal at the first level of quality and the target representation.

8. A method according to claim 1 , the method comprising:
determining the second entropy by deriving the second reconstruction data.

9. A method according to claim 1 , the method comprising:
determining that the target representation is to be obtained by the decoder device using one of the first and the second processing mode;
determining that a first further target representation of the first time sample of the signal is to be obtained by the decoder device using the other of the first and the second processing mode, the first further target representation being at a level of quality in the tiered hierarchy that is higher than the first level of quality; and
determining that a second further target representation of the first time sample of the signal is to be obtained by the decoder device using the one of the first and the second processing mode, the second further target representation being at a level of quality in the tiered hierarchy that is higher than the level of quality of the first further target representation.

10. A method according to claim 1 ,
wherein the first entropy is indicative of a coding efficiency associated with the first reconstruction data; and
wherein the second entropy is indicative of a coding efficiency associated with the second reconstruction data.

11. A method of obtaining a target representation of a first time sample of a signal, the signal being arranged in accordance with a tiered hierarchy having multiple levels of quality, the method comprising:
determining, based on received data, whether the target representation is to be obtained using a first processing mode or a second, different processing mode, the target representation being at a first level of quality in the tiered hierarchy; and
obtaining, by a decoder device, the target representation using either the first processing mode or the second processing mode based on a result of the determination,
wherein using the first processing mode to obtain the target representation comprises the decoder device processing first reconstruction data and a further representation of the first time sample of the signal that is at a second, lower level of quality in the tiered hierarchy,
wherein using the second processing mode to obtain the target representation comprises the decoder device processing second reconstruction data and a representation of a second, different time sample of the signal, and
wherein the first processing mode is to be used by the decoder device when a first entropy associated with the first reconstruction data is lower than a second entropy associated with the second reconstruction data, and the second processing mode is to be used by the decoder device when the second entropy is lower than the first entropy.

12. A method according to claim 11 , wherein the target representation and the representation of the second time sample are at a same level of quality as each other in the tiered hierarchy.

13. A method according to claim 11 , wherein one or more of the first reconstruction data and the second reconstruction data comprises intra-residual data useable to reconstruct the target representation.

14. A method according to claim 11 , wherein the using of the first processing mode by the decoder device comprises:
performing, by the decoder device, an upsampling operation on the further representation to derive a preliminary representation of the first time sample of the signal at the first level of quality; and
using, by the decoder device, the first reconstruction data and the preliminary representation to obtain the target representation.

15. A method according to claim 11 , wherein the second data comprises inter-processing data useable by the decoder device to reconstruct the target representation using the representation of the second time sample, and wherein the inter-processing data relates to the first time sample and the second time sample.

16. A method according to claim 15 , wherein the inter-processing data comprises motion compensation data relating to motion of one or more items between the representation of the second time sample and the target representation.

17. A method according to claim 11 , wherein the using of the second processing mode by the decoder

Assignees

• V Nova Int Ltd

Inventors

• Makeev Ivan
• Keszthelyi Balázs
• Ettinger Robert
• Sanna Michele
• Poularakis Stergios

Classifications

• H04N19/33

  in the spatial domain · CPC title

• H04N19/107 Primary

  between spatial and temporal predictive coding, e.g. picture refresh · CPC title

• H04N19/15 Primary

  by monitoring actual compressed data size at the memory before deciding storage at the transmission buffer · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/59

  involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution · CPC title