Back to results
US 11,778,213 B2Granted
Activation function design in neural network-based filtering process for video coding
AssigneeQUALCOMM Incorporated
InventorsHongtao Wang, Jianle Chen, Marta Karczewicz
Life of this patent
Estimating the term…
Abstract
A method of coding video data, the method comprising: reconstructing a block of the video data; and applying a Convolutional Neural Network (CNN)-based filter to the reconstructed block, wherein the CNN-based filter uses a LeakyReLU activation function.
Claims20
A method of decoding video data, the method comprising:
- obtaining one or more syntax elements from a bitstream that comprises an encoded representation of the video data;
- reconstructing a block of the video data;
- applying a Convolutional Neural Network (CNN)-based filter to the reconstructed block, wherein:
- a CNN of the CNN-based filter includes a plurality of convolutional layers,
- the CNN-based filter applies a Leaky Rectified Linear Unit (LeakyReLU) activation function to output values of the convolutional layers, and
- for each convolutional layer of the plurality of convolutional layers, the LeakyReLU activation function applied to the output values of the convolutional layer is defined as:
- f
-
- (
- y
- )
- =
- {
- Alpha
- *
- y
- ,
- y
- <
- 0
- y
- ,
- y
- ≥
- 0
- where y is an output value of the convolutional layer and Alpha is a fixed parameter;
- and
- for at least one convolutional layer of the plurality of convolutional layers, determining, based on the one or more syntax elements, a value of Alpha used in the LeakyReLU activation function applied to an output value of the at least one convolutional layer.
25 more elements
A device for decoding video data, the device comprising:
- a memory to store the video data;
- and
- one or more processors implemented in circuitry, the one or more processors configured to:
- obtain one or more syntax elements from a bitstream that comprises an encoded representation of the video data;
- reconstruct a block of the video data;
- apply a Convolutional Neural Network (CNN)-based filter to the reconstructed block, wherein:
- a CNN of the CNN-based filter includes a plurality of convolutional layers,
- the CNN-based filter applies a Leaky Rectified Linear Unit (LeakyReLU) activation function to output values of the convolutional layers, and
- for each convolutional layer of the plurality of convolutional layers, the LeakyReLU activation function applied to the output values of the convolutional layer is defined as:
- f
-
- (
- y
- )
- =
- {
- Alpha
- *
- y
- ,
- y
- <
- 0
- y
- ,
- y
- ≥
- 0
- where y is an output value of the convolutional layer and Alpha is a fixed parameter;
- and
- for at least one convolutional layer of the plurality of convolutional layers, determine, based on the one or more syntax elements, a value of Alpha used in the LeakyReLU activation function applied to an output value of the at least one convolutional layer.
28 more elements
A non-transitory computer-readable storage medium having stored thereon instructions that, when executed, cause one or more processors to:
- obtain one or more syntax elements from a bitstream that comprises an encoded representation of the video data;
- reconstruct a block of video data;
- apply Convolutional Neural Network (CNN)-based filter to the reconstructed block, wherein:
- a CNN of the CNN-based filter includes a plurality of convolutional layers,
- the CNN-based filter applies a Leaky Rectified Linear Unit (LeakyReLU) activation function to output values of the convolutional layers, and
- for each convolutional layer of the plurality of convolutional layers, the LeakyReLU activation function applied to the output values of the convolutional layer is defined as:
- f
-
- (
- y
- )
- =
- {
- Alpha
- *
- y
- ,
- y
- <
- 0
- y
- ,
- y
- ≥
- 0
- where y is an output value of the convolutional layer and Alpha is a fixed parameter;
- and
- for at least one convolutional layer of the plurality of convolutional layers, determine, based on the one or more syntax elements, a value of Alpha used in the LeakyReLU activation function applied to an output value of the at least one convolutional layer.
25 more elements
Dependencies are inferred from the claim text.