<ul><li>Amazon Nova Canvas is a generative model within Amazon Nova creative models for text-to-image generation.</li><li>To begin using Amazon Nova Canvas, set up your Amazon Bedrock account in the desired region (e.g., US East).</li><li>Amazon Nova Canvas uses diffusion to generate images, starting from noise and guided by text prompts.</li><li>Text prompts play a crucial role in guiding image generation, detailing subject, style, and composition.</li><li>Balancing positive and negative prompts helps create a defined creative space for the model.</li><li>Amazon Nova Canvas supports various image dimensions and aspect ratios for optimal results.</li><li>The Classifier-Free Guidance (CFG) scale influences prompt adherence and creative freedom in image generation.</li><li>Seed values ensure reproducibility in image generations for refining prompts and parameters.</li><li>Amazon Bedrock provides an API structure for text-to-image generation with Amazon Nova Canvas.</li><li>Cleanup steps include backing up Jupyter notebooks and shutting down SageMaker instances to avoid charges.</li></ul>

Text-to-image basics with Amazon Nova Canvas

Discover more