| Publisher | Dapper Octopus |
|---|---|
| File size | 76.17MB |
| Number of files | 81 |
| Latest version | 1 |
| Latest release date | 2020-04-14 10:47:10 |
| First release date | 2019-12-11 01:42:14 |
| Supported Unity versions | 2018.4.2 or higher |
Use learning in your core game loop:
- The algorithm updates at runtime, so teaching, praising, scolding, adapting and reinforcing can become part of the live player’s experience.
- No two game sessions will look exactly the same.
Unity integration for DyNet:
- DyNet is a dynamic neural network toolkit, similar to PyTorch.
- Developers who are experienced with neural networks will find this library familiar and well documented.
Built on highly efficient C++ code:
- This integration connects to a highly efficient compiled C++ code, enabling fast computation of neural algorithms and reducing the impact on framerate.
Neural net starter demo:
- We’ve included a demo neural network to jumpstart your machine learning projects, along with two sample projects that demonstrate it in action.
- Even if you have no experience with machine learning, this demo provides the framework for live-training a neural network during gameplay.
- The C# integration means you can convert your existing game scripts to be driven by machine learning with very little modifications to the demo AI Controller.
- Clearly commented code explains how game logic and the neural network interface.
No pre-training or big data necessary:
- Because these neural networks are generated at runtime, you can create all of your data from player input and environmental cues during play.
- Training results can be seen within minutes! Outside datasets can still be used if desired.
Use Unity’s render engine for visualization:
- Running live, you can use Unity’s UI tools to visualize the learning process.
- This enables you to find unexpected behaviors and track the algorithm’s effectiveness in an intuitive way.
- Great way for machine learning developers to showcase their results to people who are less savvy at dissecting traditional debug information.
Visit ailive.software for more information!
To enable AILIVE make sure the “Plugins” folder is in the top level (Assets) directory. You may need to restart Unity for best results.
If you are using the sample projects, make sure the “Standard Assets” is moved to the top level directory as well.
Please note that AILIVE currently only supports the Windows platform.
- The algorithm updates at runtime, so teaching, praising, scolding, adapting and reinforcing can become part of the live player’s experience.
- No two game sessions will look exactly the same.
Unity integration for DyNet:
- DyNet is a dynamic neural network toolkit, similar to PyTorch.
- Developers who are experienced with neural networks will find this library familiar and well documented.
Built on highly efficient C++ code:
- This integration connects to a highly efficient compiled C++ code, enabling fast computation of neural algorithms and reducing the impact on framerate.
Neural net starter demo:
- We’ve included a demo neural network to jumpstart your machine learning projects, along with two sample projects that demonstrate it in action.
- Even if you have no experience with machine learning, this demo provides the framework for live-training a neural network during gameplay.
- The C# integration means you can convert your existing game scripts to be driven by machine learning with very little modifications to the demo AI Controller.
- Clearly commented code explains how game logic and the neural network interface.
No pre-training or big data necessary:
- Because these neural networks are generated at runtime, you can create all of your data from player input and environmental cues during play.
- Training results can be seen within minutes! Outside datasets can still be used if desired.
Use Unity’s render engine for visualization:
- Running live, you can use Unity’s UI tools to visualize the learning process.
- This enables you to find unexpected behaviors and track the algorithm’s effectiveness in an intuitive way.
- Great way for machine learning developers to showcase their results to people who are less savvy at dissecting traditional debug information.
Visit ailive.software for more information!
To enable AILIVE make sure the “Plugins” folder is in the top level (Assets) directory. You may need to restart Unity for best results.
If you are using the sample projects, make sure the “Standard Assets” is moved to the top level directory as well.
Please note that AILIVE currently only supports the Windows platform.
