Super Resolution

This function block is used to enhance the quality of images by upsampling them using advanced deep learning techniques. It improves image resolution and quality, making it ideal for applications that require high-definition output.

📥 Inputs

Image Any Accepts any input image to which super-resolution will be applied.

📤 Outputs

Image Any This output provides the upscaled version of the input image after applying super-resolution techniques.

🕹️ Controls

SuperResolution Type A dropdown menu that allows users to select from various super-resolution models. Each model may have different performance characteristics and scaling factors.

🎨 Features

Multiple Models Available The block allows users to select from a range of super-resolution models optimized for different scaling factors and performance.

Dynamic Loading of Models The selected model is loaded dynamically upon selection, ensuring that the most suitable model is used for the operation.

Error Handling Provides meaningful error messages to handle GPU memory issues or other exceptions during the execution of the model.

📝 Usage Instructions

1. Input Image: Connect an image that you want to upscale to the Image Any input.

2. Select Model: Choose the desired super-resolution model from the SuperResolution Type dropdown.

3. Run Evaluation: Execute the block to apply the super-resolution and retrieve the enhanced image.

📊 Evaluation

Upon evaluation, this function block processes the input image through the selected super-resolution model, resulting in an enhanced image that is output through the designated socket.

💡 Tips and Tricks

Handling Large Images

If the input image is too large, consider using the Image Resizer function block to reduce the size before applying super-resolution. This can help avoid memory issues during processing.

Choosing the Right Model

Experiment with different models available in the dropdown. Some are optimized for speed, while others prioritize quality. Based on your needs, choose a model that best suits your application.

GPU Memory Management

If you encounter out-of-memory errors, try using lighter models such as FASTEST or FAST versions as they use less GPU memory compared to higher-tier models.

🛠️ Troubleshooting

GPU Out of Memory

If you receive an error about GPU memory, consider selecting a lighter model or resizing the input image to reduce resource consumption.

Model Not Loading

Ensure that the model files are correctly placed in the specified SUPER_RESOLUTION_MODEL_PATH directory. Verify that all necessary model files exist for the selected model.