OasisModel<T>
Models & Types · AiDotNet.Diffusion.Video.WorldModels
Oasis playable AI game via next-frame prediction.
Inheritance: DiffusionModelBase<T> → LatentDiffusionModelBase<T> → VideoDiffusionModelBase<T> → OasisModel<T>
For Beginners
Oasis is an AI-powered Minecraft engine that generates the next video frame based on player actions. It creates a fully playable Minecraft-like experience through real-time next-frame prediction.
How It Works
References:
- Reference: Oasis: AI-Generated Playable Minecraft (2024)
Oasis is the world’s first playable AI-generated game, reimagining Minecraft entirely through real-time AI generation. Every frame is generated by AI using next-frame prediction trained on millions of hours of Minecraft footage. The model takes player actions as input and generates the corresponding next frame in real-time.
Technical specifications:
- Architecture: Next-Frame Prediction DiT + Action Conditioning
- Latent channels: 16
- Default: 1 frames at 30 FPS
- Supports I2V: Yes | T2V: Yes | V2V: No
Example
var options = new DiffusionModelOptions<float> { LatentChannels = 16, DefaultInferenceSteps = 1 };
var actionInput = Tensor<float>.CreateRandom(new[] { 1, 16, 1, 45, 80 });
var trainX = Tensor<float>.CreateRandom(4, 8);
var trainY = Tensor<float>.CreateRandom(4, 2);
var result = new AiModelBuilder<float, Tensor<float>, Tensor<float>>()
.ConfigureModel(new OasisModel<float>(options: options))
.Build(trainX, trainY);
var nextFrame = result.Predict(actionInput);
Constructors
| Constructor | Summary |
|---|---|
OasisModel(NeuralNetworkArchitecture<T> architecture, DiffusionModelOptions<T> options, INoiseScheduler<T> scheduler, DiTNoisePredictor<T> predictor, TemporalVAE<T> temporalVAE, IConditioningModule<T> conditioner, int defaultNumFrames, int defaultFPS, int? seed) |
Initializes a new instance of OasisModel with full customization support. |
Properties
| Property | Summary |
|---|---|
IConditioningModule<T> Conditioner { get; } |
|
INoisePredictor<T> NoisePredictor { get; } |
|
IVAEModel<T> TemporalVAE { get; } |
|
IVAEModel<T> VAE { get; } |
|
bool SupportsImageToVideo { get; } |
|
bool SupportsTextToVideo { get; } |
|
bool SupportsVideoToVideo { get; } |
|
int LatentChannels { get; } |
Methods
| Method | Summary |
|---|---|
ModelMetadata<T> GetModelMetadata() |