CN105069835A - Method for achieving artificial intelligence visual realistic sense of game - Google Patents
Method for achieving artificial intelligence visual realistic sense of game Download PDFInfo
- Publication number
- CN105069835A CN105069835A CN201510426519.0A CN201510426519A CN105069835A CN 105069835 A CN105069835 A CN 105069835A CN 201510426519 A CN201510426519 A CN 201510426519A CN 105069835 A CN105069835 A CN 105069835A
- Authority
- CN
- China
- Prior art keywords
- role
- artificial intelligence
- game
- identification
- sense
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses a method for achieving artificial intelligence visual realistic sense of a game. The method comprises: creating a elliptical curve to construct a visual model of an artificial intelligence role: initializing the visual model, and intelligently computing and detecting an object and a role within a range of visibility; constructing a new small elliptical curve in the visual model by using a discrete method and allocating areas, where in each area is provided with a corresponding identification probability; detecting a self behavior and behavior changes of the object and the role in real time, and dynamically adjusting the identification probabilities. The method has beneficial effects of consuming few resources of a game program; enabling the visual model to achieve coherent a NPC (non play character) in the game and detailed cognition to a game world; enabling the visual model to provide with NPC in the game with more details and realistic sense; and enhancing the expression of the game and immersion sense of a player, and improving the interestingness and the playability of the game.
Description
Technical field
The present invention relates to a kind of method realizing game artificial intelligence sense of reality vision, belong to computer game field.
Background technology
The performance of game artificial intelligence is an important part in gaming, and it and player carry out directly mutual, are directly connected to the game experiencing of player, the lifting game quality that a good artificial intelligence is enough.
Along with developing rapidly of game, players expects that artificial intelligence role can more true to nature, perception gaming world more subtly.But as the core of artificial intelligence, the artificial intelligence role vision mode in traditional game is foolproof, usually uses the combination that sighting distance, the cone and sight line check.Although these basic artificial intelligence role vision mode functions are very efficient and be easy to write, they are concerning too transparent game player and too simple.Such as when artificial intelligence role uses discrete distance to check vision, exceeding one specifically apart from absolute blind area being had afterwards.The artificial intelligence of enemy will be handled with it after player is familiar with this leak.Such as player lures independent enemy by repeating slowly to run away forward from enemy colony again.
Fig. 1 is sighting distance, the cone, the sight line test pattern of usual artificial intelligence, and it is sighting distance, the cone and sight line respectively that the three kinds of core visual used in major part game at present calculate.In order to make these three kinds of inspections, there is efficiency usually can calculate in order.This is because range test is very efficient, and the game resource consumption of sight line test can be larger.Fig. 2 content is as follows:
The first, simple sighting distance inspection calculates.Measuring distance square instead of real distance can be more efficient because it avoid square root functions.Can the dot product of compute vector come and square the comparing of sighting distance.Why can applications distances square optimizing, is because be only concerned about relative distance, instead of definite distance.
Second, the cone is tested.First by the vectorial forward normalization of artificial intelligence role, equally by the vectorial normalization from artificial intelligence role to player, then dot product operation is carried out to them.If result is greater than 0 ° of angle, player is just in the cone at 180 ° of angles of artificial intelligence role.If result is greater than 0.5, player is just in the cone of the hexagonal angle of artificial intelligence role.As optimization, if only need the cone at test 180 ° of angles, then there is no need normalized vector, two square roots can be excluded like this and calculate.
3rd, sight line is tested, and this execution expends time in the most.Test can penetrate a ray from the level height of the eyes of artificial intelligence role to the position of player.If encountered any solid before it meets player, artificial intelligence role just cannot see player.This test can be optimized by using the bounding box of outpost of the tax office solid, instead of tests independent polygon.In order to test the object in the world, other LOD of object lowermost level and bounding box can be used.These three kinds of tests are bases of vision mode.But this simple visual test can not be that mankind's visual modeling, the particularly cone have a lot of shortcoming, shown in specific as follows well: artificial intelligence role can not see the object in its dead astern; Visual acuity is the highest and along with range attenuation at the center in the visual field, and cone model have estimated visual zone at a distance excessively, and underestimates visual zone nearby; In order to avoid the region that vision is excessive at a distance, game developer can allow sighting distance shorten untruely.
Summary of the invention
For the defect on prior art artificial intelligence visual development, to the present situation of general game artificial intelligence vision mode with cause reason analysis, summarize a kind of method realizing game artificial intelligence sense of reality vision, the artificial intelligence of game artificial intelligence role is significantly improved, the game experiencing of player is better, uses elliptic region to build artificial intelligence and too increases the interest of game and recreational simultaneously.
The technical scheme that technical solution problem of the present invention adopts comprises the following steps:
A, transfer the range information between artificial intelligence role and detected object and role in games, set up the first elliptic region, described range information is configured to the first elliptic region, and arranges that artificial intelligence role is positioned at the end points place of the transverse of the first elliptic region;
B, detect the position of object outside described artificial intelligence role and role, calculate the regional location at object and role place according to the range information of configuration, storage area positional information;
C, set up the second elliptic region, described regional location is configured to described second elliptic region, described artificial intelligence role is made to be arranged on the elliptic focus of the first elliptic region, and make the dead ahead of artificial intelligence role spread all over described second elliptic region, then on the basis of the second elliptic region, set up half-circle area further, artificial intelligence role is positioned on the center of circle of described half-circle area, and make the dead ahead of this artificial intelligence role spread all over described half-circle area;
D, adopt discrete method of testing to be divided into multiple subregion to the combination zone that described second elliptic region and described half-circle area are combined into, and every sub regions is distributed to the identification probable value being used for object outside artificial intelligence role described in identification and role;
E, when etc. the behavior state of object to be identified and role and/or described artificial intelligence role and game attributes change time, on-the-fly modify and the identification probability in the subregion of described artificial intelligence role association.Further, this game artificial intelligence sense of reality vision implementation method also comprises: use this two-dimensional elliptic curve construction artificial intelligence role vision mode to two-dimensional game or 3d gaming, if important to requirement for height in game, then can use three-dimensional elliptic curve structure artificial intelligence role vision mode; For just recognition object, role, when its distance becomes far away, the axial length of dynamic conditioning ellipse and the visual angle angle of artificial intelligence role, the vision breach that analog vision causes because of distance change.
Further, the method step A of described game artificial intelligence sense of reality vision comprises: transfer the range information between artificial intelligence role and detected object and role in games, described range information comprises position, distance length, the angle of artificial intelligence role and object to be detected and role, also comprises the cone of artificial intelligence role, sighting distance, perspective data.
Further, the method step B of described game artificial intelligence sense of reality vision comprises: according to described range information, detect the object outside described artificial intelligence role and role, calculate object outside artificial intelligence role and the role Distance geometry to the first elliptic region two focuses, if described Distance geometry is greater than the major axis of the first elliptic region, represent that object and role are outside artificial intelligence role sighting distance, now then do not carry out next step behavior enforcement, if described Distance geometry is less than or equal to the first elliptic region major axis, represent that object and role are within artificial intelligence role sighting distance, now can carry out next step behavior enforcement.
Further, the method step C of described game artificial intelligence sense of reality vision comprises: make the second described elliptic region and half-circle area form combination zone, described combination zone is the identification region of artificial Autonomous role.
Further, the method step D of described game artificial intelligence sense of reality vision comprises: according to different visual acuity Detection results, identification probable values different is accordingly distributed to every sub regions, therefore different to each zone marker identification probability, described identification probability is the probability of artificial Autonomous role identification object and role, and to represent completely by artificial intelligence role identification higher than upper limit threshold, artificial intelligence role then can not be caused to take any behavior lower than lower threshold.
Preferably, the step D of described game artificial intelligence sense of reality visible sensation method also comprises: described upper limit threshold is 100%, lower threshold is 50%, is comprise multiple threshold value section in 50% to 100% in threshold value, in described threshold value section, all can take corresponding action after artificial intelligence role identification success.
Further, the method step E of described game artificial intelligence sense of reality vision comprises: when the object detected and the behavior of role, residing region, state change, according to the difference of object and the behavior of role, residing region, state, on-the-fly modify the probability of artificial intelligence role identification object and role, when artificial intelligence role-act state changes, the identification probability of amendment artificial intelligence role to object and role also can be changed.
Further, the method step E of described game artificial intelligence sense of reality vision also comprises: for the object and the role that are in specific behavior or state, the change that artificial intelligence role can fix object and role's identification probability, and the game environment change residing for artificial intelligence role, also can on-the-fly modify the identification probability of artificial intelligence role; For being in the second elliptic region and object in artificial intelligence role front and role, can by the complete identification of artificial intelligence role, be in the second elliptic region and object at artificial intelligence role eyes rear and role, can't by complete identification, to reach sense of reality vision; For the vision of described artificial intelligence role, the decay of angle and distance and the discrimination power of artificial intelligence role that causes declines, meets the ocular rules of the mankind.
Further, the method of described game artificial intelligence sense of reality vision also comprises: for described step B, when artificial intelligence role is in inspected object and role's process, object and role cause it to have left the sighting distance of artificial intelligence because of mobile, now on-the-fly modify the artificial intelligence role visual angle size of the first elliptic region and the first elliptic region axial length length to recover the detection to object and role in time.
Further, the method of described game artificial intelligence sense of reality vision also comprises: acquiescence uses the graphics field of two-dimensional primitive structure artificial intelligence role vision in gaming, important to the requirement of height if played, then use three-dimensional pel domain construction artificial intelligence role's visual pattern region.
Beneficial effect of the present invention is: the vision mode of elliptic region structure links up to game artificial intelligence role and the detailed understanding to gaming world.A lot of very attracting characteristic is introduced into this model, and this will bring and extremely enrich interesting game play.When player understanding of the vision mode of bottom better, they just the method for design innovation can handle artificial intelligence role, and this can improve the quality of game experiencing greatly.This vision mode brings more details and the sense of reality to game NPC.It is a model very flexibly simultaneously, the vision mode that can create oneself coordinates and strengthens specific game design, the sense of reality of game artificial intelligence role and the degree of intelligence of artificial intelligence is strengthened to reach, strengthen the expressive force of game and the substitution sense of player, improve the enjoyment of game and the object of game.
Accompanying drawing explanation
Figure 1 shows that the sighting distance of usual artificial intelligence, the cone, sight line figure;
Figure 2 shows that the overview flow chart according to embodiment of the present invention;
Figure 3 shows that the artificial intelligence vision mode figure constructed according to the elliptic region of embodiment of the present invention;
Figure 4 shows that the identification probability graph of the artificial intelligence role according to embodiment of the present invention.
Embodiment
In order to make the object, technical solutions and advantages of the present invention clearly, describe the present invention below in conjunction with the drawings and specific embodiments.A kind of method realizing game artificial intelligence sense of reality vision of the present invention is applicable to the exploitation of the game such as single-play game, mobile phone games, web game, is especially used in the game of the artificial intelligence of non-player role (i.e. artificial intelligence role).
Figure 1 shows that, according to process flow diagram of the invention process, described flow process is: the vision mode using the elliptic curve structure artificial intelligence role that can customize editor; The vision mode of the artificial intelligence role of initialization elliptic curve structure, resolve and load artificial intelligence role and the object, the role distance association attributes that find, whether intelligent computation object, role enter in the visual range of artificial intelligence role; Use discrete method of testing, testing environment object, role, simulating human actual visual, at the elliptic curve that the vision mode inside establishment one of described elliptic curve structure artificial intelligence role is less, use number percent to carry out dividing by region to the object in artificial intelligence role visual range, described number percent represents that object, role are by the probability of identification; At inspected object, Jiao Seshi, artificial intelligence role region residing for object, role can carry out identification according to the probability of correspondence, when object, role are in specific position or behavior changes, dynamic amendment its by the probability size of identification, and artificial intelligence role-act is when changing, also can impact the identification probability size of artificial intelligence itself.
Figure 3 shows that the artificial intelligence vision mode constructed according to the elliptic region of embodiment of the present invention, comprise vision mode and object, person detecting computing.As shown in the figure, fBehindDist represents region after one's death, vPos is the eye position of artificial Autonomous role, vDir represents the object that artificial intelligence role looks and role region, one in elliptic region, one outside elliptic region, major axis is long is 2a, minor axis is long is 2b, and two focuses (f1 and f2) are c and-c from the distance of elliptical center, and upper figure indicates oval key property.Any point on ellipse must equal 2a to the distance sum of two focuses.In order to check that certain point is whether in ellipse, first must find the position of focus.In order to simulating human vision, one end of ellipse is placed on the eye position of NPC, planning then can specify the angle of the cone.The eyes of visual angle, NPC and elliptical center point can determine a triangle, as shown in the figure.Planning can specify a maximum sighting distance equally, and the half of sighting distance is exactly the length of NPC to elliptical center, so given θ is the half at visual angle, and a is the half of sighting distance, first provides c, the formula between θ and a.
Following formula is obtained by substituting into:
Just precomputation can be carried out when initialization.Need the position determining focus.The eyes of artificial intelligence role are at vPos point, and the direction seen is vDir point, and formula is as follows:
If want to allow oval fBehindDist place beginning after artificial intelligence role, such NPC can perception with the people of his, final formula is as follows:
In order to save unnecessary subtraction operation, fBehindDist just can cut when initialization from a.
Whether enter in the visual field of artificial intelligence role to detect an object, only needing to calculate object to the distance of two focuses, and check and be consequently noly less than maximum viewing distance 2a if being added.So for each object, each artificial intelligence, each role only needs to do twice distance and checks.
Figure 4 shows that the visual acuity of first carrying out and color detection are the highest in optic centre, and decline rapidly at periphery according to identification probability imitation human vision figure of the invention process.When peripheral vision is weak time, motion detection becomes sharper.Use discrete test to create vision mode, score to give the object in visual range.In the drawings, number percent represents that certain objects is by the probability of identification.The object of optic centre is by fully identification, and at nearly periphery, middle periphery, there is lower probability in periphery far away and 6th sense region after one's death.A key property of this model is the bonus point that the object of movement obtains 50%, as the special vision to motion.Concerning game, walk and run may trigger 50% bonus point, but move under water or creep can not.If pretend or be hidden in shade to serve very important effect in gaming, just need to combine reference contrast with the profile exposed to weaken identification.Such as when player is couchant in dark corners, their profile will be less and be difficult to see clearly, and result just needs 100% ground to weaken their identification.Be just that NPC directly look at player, perhaps NPC only can stare at several seconds and then moves on, because it can not 100% ground identification object.If use this nuance, then add a less ellipse, identification can be unconditionally 100% wherein, and without end out line and contrast.The number percent of probability can be explained by reasonable manner in game design.Method is that artificial intelligence role can show different behaviors in different threshold range.Such as under the probability of 100%, object is by complete identification, and artificial intelligence role can shoot object.Under the probability of 80% or higher, artificial intelligence role understands rotary head and starts to process object identification.Under the probability of 50% or higher, NPC has probability meeting rotary head.Any probability below 50% is all not enough to stimulate NPC to take any behavior.
The above, just preferred embodiment of the present invention, the present invention is not limited to above-mentioned embodiment, as long as it reaches technique effect of the present invention with identical means, all should belong to protection scope of the present invention.In protection scope of the present invention, its technical scheme and/or embodiment can have various different modifications and variations.
Claims (10)
1. realize a method for game artificial intelligence sense of reality vision, it is characterized in that, the method comprises:
A, transfer the range information between artificial intelligence role and detected object and role in games, set up the first elliptic region, described range information is configured to the first elliptic region, and arranges that artificial intelligence role is positioned at the end points place of the transverse of the first elliptic region;
B, detect the position of object outside described artificial intelligence role and role, calculate the regional location at object and role place according to the range information of configuration, storage area positional information;
C, set up the second elliptic region, described regional location is configured to described second elliptic region, described artificial intelligence role is made to be arranged on the elliptic focus of the first elliptic region, and make the dead ahead of artificial intelligence role spread all over described second elliptic region, then on the basis of the second elliptic region, set up half-circle area further, artificial intelligence role is positioned on the center of circle of described half-circle area, and make the dead ahead of this artificial intelligence role spread all over described half-circle area;
D, adopt discrete method of testing to be divided into multiple subregion to the combination zone that described second elliptic region and described half-circle area are combined into, and every sub regions is distributed to the identification probable value being used for object outside artificial intelligence role described in identification and role;
E, when etc. the behavior state of object to be identified and role and/or described artificial intelligence role and game attributes change time, on-the-fly modify and the identification probability in the subregion of described artificial intelligence role association.
2. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, described steps A comprises:
Transfer the range information between artificial intelligence role and detected object and role in games, described range information comprises position, distance length, the angle of artificial intelligence role and object to be detected and role, also comprises the cone of artificial intelligence role, sighting distance, perspective data.
3. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, described step B comprises:
According to described range information, detect the object outside described artificial intelligence role and role, calculate object outside artificial intelligence role and the role Distance geometry to the first elliptic region two focuses, if described Distance geometry is greater than the major axis of the first elliptic region, represent that object and role are outside artificial intelligence role sighting distance, now then do not carry out next step behavior enforcement, if described Distance geometry is less than or equal to the first elliptic region major axis, represent that object and role are within artificial intelligence role sighting distance, now can carry out next step behavior enforcement.
4. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, described step C comprises:
Make the second described elliptic region and half-circle area form combination zone, described combination zone is the identification region of artificial Autonomous role.
5. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, described step D comprises:
According to different visual acuity Detection results, identification probable values different is accordingly distributed to every sub regions, therefore different to each zone marker identification probability, described identification probability is the probability of artificial Autonomous role identification object and role, and to represent completely by artificial intelligence role identification higher than upper limit threshold, artificial intelligence role then can not be caused to take any behavior lower than lower threshold.
6. a kind of method realizing game artificial intelligence sense of reality vision according to claim 5, it is characterized in that, described step D also comprises:
Described upper limit threshold is 100%, and lower threshold is 50%, is comprise multiple threshold value section in 50% to 100% in threshold value, in described threshold value section, all can take corresponding action after artificial intelligence role identification success.
7. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, described step e comprises:
When the object detected and the behavior of role, residing region, state change, according to the difference of object and the behavior of role, residing region, state, on-the-fly modify the probability of artificial intelligence role identification object and role, when artificial intelligence role-act state changes, the identification probability of amendment artificial intelligence role to object and role also can be changed.
8. a kind of method realizing game artificial intelligence sense of reality vision according to claim 7, it is characterized in that, described step e also comprises:
For the object and the role that are in specific behavior or state, the change that artificial intelligence role can fix object and role's identification probability, and the game environment change residing for artificial intelligence role, also can on-the-fly modify the identification probability of artificial intelligence role;
For being in the second elliptic region and object in artificial intelligence role front and role, can by the complete identification of artificial intelligence role, be in the second elliptic region and object at artificial intelligence role eyes rear and role, can't by complete identification, to reach sense of reality vision;
For the vision of described artificial intelligence role, the decay of angle and distance and the discrimination power of artificial intelligence role that causes declines, meets the ocular rules of the mankind.
9. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, the method also comprises:
For described step B, when artificial intelligence role is in inspected object and role's process, object and role cause it to have left the sighting distance of artificial intelligence because of mobile, now on-the-fly modify the artificial intelligence role visual angle size of the first elliptic region and the first elliptic region axial length length to recover the detection to object and role in time.
10. a kind of method realizing game artificial intelligence sense of reality vision according to claim 1, it is characterized in that, the method also comprises:
Acquiescence uses the graphics field of two-dimensional primitive structure artificial intelligence role vision in gaming, if the requirement of game to height is important, then uses three-dimensional pel domain construction artificial intelligence role's visual pattern region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510426519.0A CN105069835B (en) | 2015-07-20 | 2015-07-20 | A kind of method for realizing game artificial intelligence sense of reality vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510426519.0A CN105069835B (en) | 2015-07-20 | 2015-07-20 | A kind of method for realizing game artificial intelligence sense of reality vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105069835A true CN105069835A (en) | 2015-11-18 |
| CN105069835B CN105069835B (en) | 2018-02-13 |
Family
ID=54499191
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510426519.0A Active CN105069835B (en) | 2015-07-20 | 2015-07-20 | A kind of method for realizing game artificial intelligence sense of reality vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105069835B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107256174A (en) * | 2017-05-27 | 2017-10-17 | 武汉秀宝软件有限公司 | The implementation method and device of artificial intelligence |
| CN110898433A (en) * | 2019-11-28 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Virtual object control method and device, electronic equipment and storage medium |
| CN110935170A (en) * | 2019-10-29 | 2020-03-31 | 广州西山居世游网络科技有限公司 | Game art resource distribution lookup method and system |
| CN111450533A (en) * | 2020-03-31 | 2020-07-28 | 腾讯科技（深圳）有限公司 | Virtual object control method, device, terminal and storage medium in virtual scene |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006320384A (en) * | 2005-05-17 | 2006-11-30 | Heiwa Corp | Game machine |
| JP2007301137A (en) * | 2006-05-11 | 2007-11-22 | Aruze Corp | Slot machine |
| CN101639927A (en) * | 2008-07-31 | 2010-02-03 | 国际商业机器公司 | Method and system for adjusting virtual display device in virtual world |
| CN103155002A (en) * | 2010-10-06 | 2013-06-12 | 阿尔卡特朗讯 | Method and arrangement for identifying virtual visual information in images |
-
2015
- 2015-07-20 CN CN201510426519.0A patent/CN105069835B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006320384A (en) * | 2005-05-17 | 2006-11-30 | Heiwa Corp | Game machine |
| JP2007301137A (en) * | 2006-05-11 | 2007-11-22 | Aruze Corp | Slot machine |
| CN101639927A (en) * | 2008-07-31 | 2010-02-03 | 国际商业机器公司 | Method and system for adjusting virtual display device in virtual world |
| CN103155002A (en) * | 2010-10-06 | 2013-06-12 | 阿尔卡特朗讯 | Method and arrangement for identifying virtual visual information in images |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107256174A (en) * | 2017-05-27 | 2017-10-17 | 武汉秀宝软件有限公司 | The implementation method and device of artificial intelligence |
| CN110935170A (en) * | 2019-10-29 | 2020-03-31 | 广州西山居世游网络科技有限公司 | Game art resource distribution lookup method and system |
| CN110935170B (en) * | 2019-10-29 | 2023-09-15 | 广州西山居网络科技有限公司 | Game art resource distribution consulting method and system |
| CN110898433A (en) * | 2019-11-28 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Virtual object control method and device, electronic equipment and storage medium |
| CN111450533A (en) * | 2020-03-31 | 2020-07-28 | 腾讯科技（深圳）有限公司 | Virtual object control method, device, terminal and storage medium in virtual scene |
| CN111450533B (en) * | 2020-03-31 | 2022-06-24 | 腾讯科技（深圳）有限公司 | Virtual object control method, device, terminal and storage medium in virtual scene |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105069835B (en) | 2018-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Beattie et al. | Deepmind lab | |
| CN106267822A (en) | The method of testing of game performance and device | |
| CN111111220B (en) | Self-chess-playing model training method and device for multiplayer battle game and computer equipment | |
| CN105069835A (en) | Method for achieving artificial intelligence visual realistic sense of game | |
| CN108211362A (en) | A kind of non-player role fight policy learning method based on depth Q learning networks | |
| CN108421257A (en) | Determination method, apparatus, storage medium and the electronic device of invisible element | |
| CN106362396A (en) | Game performance test method and apparatus | |
| CN109641150A (en) | Method for creating virtual objects | |
| Lazaridis et al. | Hitboxes: A survey about collision detection in video games | |
| Tremblay et al. | Adaptive companions in FPS games. | |
| Youssef et al. | Building your kingdom imitation learning for a custom gameplay using unity ml-agents | |
| Howlett et al. | Evolving pixel shaders for the prototype video game subversion | |
| Svensson et al. | Influence Map-based controllers for Ms. PacMan and the ghosts | |
| US20210118230A1 (en) | Navigation system and method | |
| KR101692539B1 (en) | NPC decision making modelling method in dynamic terrain analysis by prospect theory | |
| CN115175750A (en) | AI-based game application content generation | |
| Tomilin et al. | LevDoom: A benchmark for generalization on level difficulty in reinforcement learning | |
| Gan et al. | Open: An open-ended physics environment for learning without a task | |
| Laird | Design goals for autonomous synthetic characters | |
| Lin et al. | Design and implementation of an OpenGL based 3D first person shooting game | |
| Ge et al. | 2D Flying Shooting Game: Game Art using Circular Collision Detection and Target Tracking: Game Art using Circular Collision Detection and Target Tracking | |
| Pasztor et al. | Online RPG Environment for Reinforcement Learning | |
| CN117009206A (en) | Method, device, equipment and storage medium for predicting hotspot aggregation position | |
| GRÖNQVIST et al. | Baa! A procedural game based on real-time flocking behaviour | |
| Ge | Physical Reasoning for Intelligent Agent in Simulated Environments |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: 519000 Room 102, 202, 302 and 402, No. 325, Qiandao Ring Road, Tangjiawan Town, high tech Zone, Zhuhai City, Guangdong Province, Room 102 and 202, No. 327 and Room 302, No. 329 Patentee after: Zhuhai Jinshan Digital Network Technology Co.,Ltd. Patentee after: CHENGDU XISHANJU INTERACTIVE ENTERTAINMENT TECHNOLOGY Co.,Ltd. Address before: 519000 Jinshan software building, 8 Lanshan lane, Jida Jingshan Hill Road, Zhuhai, Guangdong Patentee before: ZHUHAI KINGSOFT ONLINE GAME TECHNOLOGY Co.,Ltd. Patentee before: CHENGDU XISHANJU INTERACTIVE ENTERTAINMENT TECHNOLOGY Co.,Ltd. |
|
| CP03 | Change of name, title or address |