Implementing machine learning models inside a game engine can be challenging. EasyComputerVision simplifies this process by combining Unity's Sentis with a no-code interface. This allows you to run computer vision models without writing code or having knowledge of machine learning. The system is fully offline, operating on a choice of CPU or GPU, without relying on costly cloud services.
You can import your model, use your webcam or in-game camera as input, and make predictions. We offer easy-to-follow tutorials and speedy customer service for any queries you may have.
EasyComputerVision is designed for image classification tasks and object detection. Future updates will provide tested compatibility for more computer vision types, such as image segmentation and depth estimation.
You can view our documentation, roadmap, tutorials, free models, and make feature requests on our GitHub page.
EasyComputerVision is compatible with models in the .onnx format. You can find thousands of freely available models to choose from, and numerous models are already available in the .onnx format in the ONNX Model Zoo.
If you have any questions, including those regarding model compatibility, please contact us via your preferred method shown on our publisher page.
Key features:
- Image classification
- Object detection
- Use in-game camera as input
- Use webcam as input
- Automatically creates a render texture using the model's input dimensions
- Inference on update
- Inference on interval
- Inference on interval using scheduled execution
- Inference on key press
- Inference on key press using scheduled execution
- Use a range of engine backend types compatible with GPU and CPU
- Display the input texture as image on a canvas
- View the output of object detection models in-game
- Modify the bounding box and text color how you want
- View a summary of the model for convenience
- Scheduled execution can be fine-tuned, with the number of model layers processed per frame configurable, enabling greater control of the inference overhead
