Adaptive loop filter classifiers
Example methods and devices are disclosed relating to the use of adaptive loop filter classifiers. An example method includes determining a sum of absolute values of one or more reconstructed residual sample values within a window associated with a block of the video data. The example method includes determining a class index based at least in part on the sum of absolute values. The example method includes determining an adaptive loop filter based on the class index. The example method includes applying the adaptive loop filter to a current to-be-filtered sample of the block. The example method includes decoding the block based on the application of the adaptive loop filter.
1 . A method of decoding video data, the method comprising:
determining a sum of absolute values of one or more reconstructed residual sample values within a window associated with a block of the video data;
determining a class index based at least in part on the sum of absolute values;
determining an adaptive loop filter based on the class index;
applying the adaptive loop filter to a current to-be-filtered sample of the block; and
decoding the block based on the application of the adaptive loop filter, wherein determining the sum of absolute values and determining the class index comprises applying a sum of absolute values classifier to the one or more reconstructed residual sample values within the window.
2 . The method of claim 1 , wherein the adaptive loop filter comprises a fixed adaptive loop filter or an adaptive loop filter signaled in a bitstream.
3 . The method of claim 1 , wherein the class index is restricted to be in a range of 0 to a positive integer value.
4 . The method of claim 1 , wherein the class index is further based on a scaling factor.
5 . The method of claim 4 , wherein the scaling factor is based on at least one of a quantization parameter, a quantization step size, a prediction mode, or a sample bit depth.
6 . The method of claim 4 , wherein the class index is based on the sum of the absolute values divided by the scaling factor.
7 . The method of claim 1 , further comprising refraining from applying a geometric transpose to the adaptive loop filter.
8 . The method of claim 1 , further comprising applying a geometric transpose to the adaptive loop filter.
9 . The method of claim 1 , further comprising parsing, from a bitstream, an identification of a classifier indicative of a sum of absolute values classifier and an indication that the reconstructed residual sample values are to be input to the classifier.
10 . The method of claim 1 , wherein the current to-be-filtered sample is a first current to-be-filtered sample, the window is a first window, the class index is a first class index, the adaptive loop filter is a first adaptive loop filter, and the block is a first block, the method further comprising:
applying a classifier to one or more inputs associated with a second window associated with a second block of the video data to determine a second class index;
applying a second adaptive loop filter to a second current to-be-filtered sample based on the second class index; and
decoding the second block based on the application of the second adaptive loop filter.
11 . The method of claim 10 , wherein the one or more inputs associated with the second window comprise predictors, de-quantized transform coefficients, reconstructed residual sample values, absolute values of reconstructed residual sample values, or samples before deblocking filtering.
12 . The method of claim 10 , wherein the classifier comprises a Laplacian-based classifier, a band-based classifier, a sum of absolute values classifier, a sum of square values classifier, a variance classifier, an entropy classifier, or a total variation classifier.
13 . A device for decoding video data, the device comprising:
one or more memories configured to store the video data; and
one or more processors implemented in circuitry and communicatively coupled to the one or more memories, the one or more processors being configured to:
determine a sum of absolute values of one or more reconstructed residual sample values within a window associated with a block of the video data;
determine a class index based at least in part on the sum of absolute values;
determine an adaptive loop filter based on the class index;
apply the adaptive loop filter to a current to-be-filtered sample of the block; and
decode the block based on the application of the adaptive loop filter, wherein as part of determining the sum of absolute values and determining the class index, the one or more processors are configured to apply a sum of absolute values classifier to the one or more reconstructed residual sample values within the window.
14 . The device of claim 13 , wherein the adaptive loop filter comprises a fixed adaptive loop filter or an adaptive loop filter signaled in a bitstream.
15 . The device of claim 13 , wherein the class index is restricted to be in a range of 0 to a positive integer value.
16 . The device of claim 13 , wherein the class index is further based on a scaling factor.
17 . The device of claim 16 , wherein the scaling factor is based on at least one of a quantization parameter, a quantization step size, a prediction mode, or a sample bit depth.
18 . The device of claim 16 , wherein the class index is based on the sum of the absolute values divided by the scaling factor.
19 . The device of claim 13 , wherein the one or more processors are further configured to refrain from applying a geometric transpose to the adaptive loop filter.
20 . The device of claim 13 , wherein the one or more processors are further configured to apply a geometric transpose to the adaptive loop filter.
21 . The device of claim 13 , wherein the one or more processors are further configured to parse, from a bitstream, an identification of a classifier indicative of a sum of absolute values classifier and an indication that the reconstructed residual sample values are to be input to the classifier.
22 . The device of claim 13 , wherein the current to-be-filtered sample is a first current to-be-filtered sample, the window is a first window, the class index is a first class index, the adaptive loop filter is a first adaptive loop filter, and the block is a first block, and wherein the one or more processors are further configured to:
apply a classifier to one or more inputs associated with a second window associated with a second block of the video data to determine a second class index;
apply a second adaptive loop filter to a second current to-be-filtered sample based on the second class index; and
decode the second block based on the application of the second adaptive loop filter.
23 . The device of claim 22 , wherein the one or more inputs associated with the second window comprise predictors, de-quantized transform coefficients, reconstructed residual sample values, absolute values of reconstructed residual sample values, or samples before deblocking filtering.
24 . The device of claim 22 , wherein the classifier comprises a Laplacian-based classifier, a band-based classifier, a sum of absolute values classifier, a sum of square values classifier, a variance classifier, an entropy classifier, or a total variation classifier.
25 . The device of claim 13 , further comprising a display configured to display decoded video data.
26 . The device of claim 13 , further comprising one or more of a camera, a computer, a mobile device, a broadcast receiver device, or a set.
27 . A method of encoding video data, the method comprising:
determining a sum of absolute values of one or more reconstructed residual sample values within a window associated with a block of the video data;
determining a class index based at least in part on the sum of absolute values;
determining an adaptive loop filter based on the class index;
applying the adaptive loop filter to a current to-be-filtered sample of the block; and
encoding the block based on the application of the adaptive loop filter, wherein as part of determining the sum of absolute values and determining the class index, the one or more processors are configured to apply a sum of absolute values classifier to the one or more reconstructed residual sample values within the window.
28 . A device for encoding video data, the device comprising:
one or more memories configured to store the video data; and
one or more processors implemented in circuitry and communicatively coupled to the one or more memories, the one or more processors being configured to:
determine a sum of absolute values of one or more reconstructed residual sample values within a window associated with a block of the video data;
determine a class index based at least in part on the sum of absolute values;
determine an adaptive loop filter based on the class index;
apply the adaptive loop filter to a current to-be-filtered sample of the block; and
encode the block based on the application of the adaptive loop filter, wherein as part of determining the sum of absolute values and determining the class index, the one or more processors are configured to apply a sum of absolute values classifier to the one or more reconstructed residual sample values within the window.
29 . The method of claim 1 , further comprising reconstructing the current-to-be-filtered sample of the block prior to applying the adaptive loop filter.
30 . The device of claim 13 , wherein the one or more processors is configured to reconstruct the current-to-be-filtered sample of the block prior to applying the adaptive loop filter.