Method of providing one or more sets of graphics parameters and computer executing a program implementing method of providing one or more sets of graphics parameters

Patent metadata

Field	Value
Publication number	US-2022347583-A1
Application number	US-201917635800-A
Country	US
Kind code	A1
Filing date	Sep 27, 2019
Priority date	Sep 27, 2019
Publication date	Nov 3, 2022
Grant date	—

Abstract

A method of providing optimized settings of graphics parameters for a computer gaming application includes: consolidating data related to settings of graphics parameters for different computer hardware equipment and respective performance values; training a machine learning model based on the consolidated data; determining a weight for each setting of a graphics parameter, by the trained machine learning model; for each set of graphics parameters, predicting a performance value achievable by the computer gaining application when it is executed on a specific type of computer, by the trained machine learning model; assigning a priority value to each graphics parameter based on its contribution to the performance value; choosing or generating at least one set of graphics parameters providing an optimized performance value, based on the predicted performance value associated with each set of graphics parameters and the determined weight for each graphics parameter and/or based on the assigned priority value.

First claim

1 . A method of providing optimized settings of graphics parameters for a computer gaming application on a specific type of computer, the method comprising:
consolidating data related to settings of graphics parameters for different computer hardware equipment and a respective performance value corresponding to each set of graphics parameters of a plurality of sets of graphics parameters, wherein each set of graphics parameters is corresponding to a respective computer hardware equipment, and wherein the respective performance value represents a level of performance of the computer gaming application when being executed under involvement of a respective computer hardware equipment; training a machine learning model based on the plurality of sets of graphics parameters and the corresponding performance value in relation to the respective computer hardware equipment; determining a weight for each setting of a graphics parameter by the trained machine learning model, wherein the weight is indicative of an impact of the setting of the respective graphics parameter on the performance value; for each set of graphics parameters, predicting a performance value achievable by the computer gaming application when it is executed on a specific type of computer including one or more of said computer hardware equipment, by the trained machine learning model; assigning a priority value to each graphics parameter based on its contribution to the performance value of the set of graphics parameters; and choosing or generating at least one set of graphics parameters providing an optimized performance value, based on the predicted performance value associated with each set of graphics parameters and further based on the determined weight for each graphics parameter and/or based on the assigned priority value.

2 . The method of claim 1 , wherein the priority value is further dependent on user preferences.

3 . The method of claim 1 , wherein the performance value comprises a frame rate.

4 . The method of claim 1 , further comprising filtering the consolidated data,
wherein filtering the consolidated data comprises:
determining types of information that do not impact output of the machine learning model; and
removing the determined types of information.

5 . The method of claim 4 , wherein filtering the input data further comprises:
removing data in the consolidated data that correspond to performance values that fall outside a predetermined range of target performance values.

6 . The method of claim 4 , wherein filtering the input data further comprises:
categorising the input data according to at least one of features of computer hardware and application; and in each category, removing data that deviate from a normal distribution of the category.

7 . The method of claim 4 , wherein filtering the input data further comprises:
determining a linear relationship between the determined weights and the performance values; and removing data from the input data based on the determined linear relationship.

8 . The method of claim 1 , wherein generating at least one set of graphics parameters for optimizing the performance value comprises normalizing and linearizing all possible combinations of settings to generate a linear regression model.

9 . The method of claim 1 , wherein consolidating the data comprises crowd-sourcing the data.

10 . The method of claim 1 , wherein determining the weight for each setting of a graphics parameter comprises using permutation feature importance determination to compute importance score for each setting of the graphics parameter, wherein the weight is indicative of the importance score.

11 . The method of claim 1 , wherein each set of the chosen or generated at least one set of graphics parameters correspond to a gaming mode of a plurality of gaming modes, wherein the plurality of gaming modes comprises a first mode optimized for graphics quality, a second mode optimized for speed of the computer gaming application, and a third mode that balances graphics quality with speed of the computer gaming application.

12 . The method of claim 11 , further comprising:
receiving a user input indicating one gaming mode from the plurality of gaming modes; and configuring the computer gaming application according to the graphics parameters corresponding to the gaming mode indicated in the user input.

13 . The method of claim 11 , further comprising:
displaying a graphical user interface that displays visual representations of the plurality of gaming modes and their corresponding predicted performance values.

14 . A computer executing a program implementing a method of providing optimized settings of graphics parameters for a computer gaming application on a specific type of computer, the method comprising:
consolidating data related to settings of graphics parameters for different computer hardware equipment and a respective performance value corresponding to each set of graphics parameters of a plurality of sets of graphics parameters, wherein each set of graphics parameters is corresponding to a respective computer hardware equipment, and wherein the respective performance value represents a level of performance of the computer gaming application when being executed under involvement of a respective computer hardware equipment; training a machine learning model based on the plurality of sets of graphics parameters and the corresponding performance value in relation to the respective computer hardware equipment; determining a weight for each setting of a graphics parameter by the trained machine learning model, wherein the weight is indicative of an impact of the setting of the respective graphics parameter on the performance value; for each set of graphics parameters, predicting a performance value achievable by the computer gaming application when it is executed on a specific type of computer including one or more of said computer hardware equipment, by the trained machine learning model; assigning a priority value to each graphics parameter based on its contribution to the performance value of the set of graphics parameters; and choosing or generating at least one set of graphics parameters providing an optimized performance value, based on the predicted performance value associated with each set of graphics parameters and further based on the determined weight for each graphics parameter and/or based on the assigned priority value.

15 . (canceled)

16 . The computer of claim 14 , wherein the priority value is further dependent on user preferences.

17 . The computer of claim 14 , wherein the performance value comprises a frame rate.

18 . The computer of claim 14 , further comprising filtering the consolidated data,
wherein filtering the consolidated data comprises:
determining types of information that do not impact output of the machine learning model; and
removing the determined types of information.

19 . The computer of claim 18 , wherein filtering the input data further comprises:
removing data in the consolidated data that correspond to performance values that fall outside a predetermined range of target performance values.

20 . The computer of claim 18 , wherein filtering the input data further comprises:
categorising the input data according to at least one of features of computer hardware and application; and in each category, removing data that deviate from a normal distribution of the category.

21 . A non-transitory computer-reada

Assignees

• Razer Asia Pacific Pte Ltd

Inventors

• He Yun
• Zhu Xu
• Chen Dejun

Classifications

• A63F13/77 Primary

  involving data related to game devices or game servers, e.g. configuration data, software version or amount of memory · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F3/04847

  Interaction techniques to control parameter settings, e.g. interaction with sliders or dials · CPC title

• G09G2360/08

  Power processing, i.e. workload management for processors involved in display operations, such as CPUs or GPUs · CPC title

• G09G5/363

  Graphics controllers · CPC title