Support NeoGAF

[IbizaPocholo]

Vid2Player3D

We present a system that learns diverse, physically simulated tennis skills from large-scale demonstrations of tennis play harvested from broadcast videos. Our approach is built upon hierarchical models, combining a low-level imitation policy and a high-level motion planning policy to steer the...

research.nvidia.com

• A system that learns diverse, physically simulated tennis skills from large-scale demonstrations of tennis play harvested from broadcast videos is presented.
• The approach combines a low-level imitation policy and a high-level motion planning policy to steer the character in a motion embedding learned from broadcast videos.
• The system can learn complex tennis shotmaking skills and realistically chain together multiple shots into extended rallies, using only simple rewards and without explicit annotations of stroke types.
• The estimated motion from broadcast videos is corrected with physics-based imitation, and a hybrid control policy is used to override erroneous aspects of the learned motion embedding with corrections predicted by the high-level policy.
• The system can produce controllers for physically-simulated tennis players that can hit the incoming ball to target positions accurately using a diverse array of strokes, spins, and playing styles.
• Two physically simulated characters playing extended tennis rallies with simulated racket and ball dynamics can be synthesized by the system.

The paper presents a system that enables physically simulated characters to learn complex tennis skills from broadcast videos. The system uses a combination of kinematic motion estimation, low-level imitation policy training, correction of motion artifacts, and high-level motion planning policy training to generate natural and precise tennis motions. The resulting simulated characters can hit incoming tennis balls with diverse and complex skills, and adopt different playing styles by learning from different player motion data. The system’s hybrid control structure and motion embedding help smooth emotions and mitigate jittering artifacts, while the correction of wrist motion allows for precise control of the tennis racket. The system can also generate novel animations of tennis rallies between different players.

 

[k_trout]

at some point I'm going to able to generate the game I want by typing in some shit

-I want a sexy tennis based rpg where aliens invade during my wimbeldon finals

 

[HL3.exe]

Great! Thanks for sharing this paper.

I've been waiting years for physically stimulated characters and game-logic to be more normalized in every modern game. This will solve a lot of clipping and collision issues and makes games more reactive and believable when interacting with them. (instead of what we have now, being mostly only visually believable). Creating a large possibly space for emergent situations.

Sidenote: I feel like Euphoria (mostly only in Rockstar's Games) was a big step up in this case back in the 00's, but was still heavy reliant on blended canned animation. After that leap this tech or th drive to make games more physically believable stagnated heavy in favor of graphics rendering. So here is to hoping more neural network learning with a focus on physical simulation make this more normalized and doable on consumer hardware (and modern game design).

 

[Davevil]

still prefer

 

[feynoob]

at some point I'm going to able to generate the game I want by typing in some shit

-I want a sexy tennis based rpg where aliens invade during my wimbeldon finals

If you want to make your own game, here are some steps.

Spoiler: Instructions

Certainly! Here's a brief breakdown of what your game might look like based on the concept you provided:

Game Title: Sexy Tennis Invasion

Game Genre: RPG / Sports

Gameplay: The game would start with the player character (a professional tennis player) competing in the Wimbledon finals against a rival player. During the match, aliens suddenly invade the stadium and disrupt the game. The player must then navigate through the chaos of the alien invasion, all while trying to protect themselves and their loved ones. The game would feature a mix of tennis gameplay and RPG elements, such as character customization, dialogue choices, and decision-making that affects the story's outcome.

Game Mechanics: The tennis gameplay would involve standard tennis rules with a few added twists, such as power-ups that grant the player temporary boosts or allow them to perform special moves. The RPG mechanics would include leveling up the player's abilities, making dialogue choices that affect the game's story, and exploring the game world to find new items and secrets.

Art Style: The game could have a cartoonish or anime-inspired art style, with vibrant colors and exaggerated character designs.

Platforms: The game could be released on PC and console platforms, such as Steam, PlayStation, and Xbox.

Of course, this is just a brief overview, and there are many more details to consider when creating a game. I hope this gives you a better idea of what your game might look like!

Spoiler: Process

Writing the code for a game can be a complex process, but here are some tips to get you started:

1. Choose a game engine: Select a game engine that fits your needs and level of expertise. Game engines like Unity, Unreal Engine, and Godot are popular options that offer a lot of resources and support.

2. Set up your development environment: Install the necessary software and tools, and configure your workspace to suit your workflow.

3. Plan your game mechanics: Think about what features and mechanics you want to include in your game, such as player movement, enemy AI, physics, collision detection, and so on.

4. Write your code: Start coding your game mechanics using the language supported by your game engine. For example, if you're using Unity, you might use C#.

5. Test and debug: Test your game regularly and debug any issues that arise. This will involve using debugging tools, reading logs, and other troubleshooting methods.

6. Optimize: Optimize your code to improve your game's performance and make sure it runs smoothly.

7. Continue to iterate: Keep working on your game, adding new features, fixing bugs, and polishing your gameplay until you have a complete, polished game.

Remember that game development can be a challenging and time-consuming process, but with dedication and hard work, you can create an awesome game. Don't hesitate to seek out resources, tutorials, and forums for help along the way. Good luck!

You can do the rest with chatgpt. Hopefully you can make a mini game out of this.

 

[RoboFu]

I may be crazy , but it seems all these ai technologies Nvidia are making are all point to an ai dating game. Is that their final goal? Lol

 

[Buggy Loop]

I may be crazy , but it seems all these ai technologies Nvidia are making are all point to an ai dating game. Is that their final goal? Lol

It’ll be what you want it to be

 

[hyperbertha]

Great! Thanks for sharing this paper.

I've been waiting years for physically stimulated characters and game-logic to be more normalized in every modern game. This will solve a lot of clipping and collision issues and makes games more reactive and believable when interacting with them. (instead of what we have now, being mostly only visually believable). Creating a large possibly space for emergent situations.

Mmmmmmmmmmm

 

[draliko]

This is the only thing that will save game Dev in the future, eliminate tedious costly repetitive stuff like mocap and generate ai physically correct movements, the search for graphics can't go on like it is now... Too resource intensive for trivial details, the budget need to go to gameplay Dev time , let the is help the artists with their vision.