Text-to-image model
A text-to-image model is a generative machine-learning model that produces images conditioned on textual descriptions or prompts.
Build a Text-to-Image Generator (from Scratch)
A text-to-image model is a machine-learning model that generates images from textual descriptions.
- How it is used
- Turns written prompts into images for visual creation, prototyping, and text-guided editing.
- How deep it goes
- Works through transformer token prediction and diffusion pipelines, including component-level implementation; uses pretrained parameters where full training is costly.
- Examples
- Traces a top-hatted panda through successive image tokens and a motorcycle-riding banana through latent diffusion.
- What you can do after
- Build it from scratch as a compact transformer-based generator, or apply a pretrained latent diffusion model to prompts.
Read how Build a Text-to-Image Generator (from Scratch) explains text-to-image model →
Learn Generative AI with PyTorch
The generative task of producing an image conditioned on a text prompt.
- How it is used
- Uses prompt-conditioned image generation to connect multimodal Transformers with diffusion and a DALL-E 2 application.
- How deep it goes
- Explains conditioning through embeddings and iterative denoising, then demonstrates an API call rather than building a text-to-image model from scratch.
- Examples
- Contrasts denoised flower images with DALL-E 2 prompts for an astronaut riding a unicorn and a cat working on a computer.
- What you can do after
- Apply it via the OpenAI Python client to generate and view a DALL-E 2 image from a chosen prompt.
Read how Learn Generative AI with PyTorch explains text-to-image model →
Introduction to Generative AI, Second Edition
A text-to-image model is a generative model that creates images from textual descriptions.
- How it is used
- Frames questions about artists’ copyrighted training images and copyright protection for generated images.
- How deep it goes
- Explains legal distinctions around generated images and training data without describing model internals.
- Examples
- Shows an illustration prompted with an artist’s webcomic and a meme restyled with GPT-4o.
- What you can do after
- Evaluate/understand why training-data rights and copyright in generated images require separate consideration.
Read how Introduction to Generative AI, Second Edition explains text-to-image model →
Deep Learning with R, Third Edition
A text-to-image model is a generative model that produces images corresponding to textual descriptions.
- How it is used
- Generates images from natural-language descriptions for creative visual exploration.
- How deep it goes
- Explains latent spaces, text conditioning, denoising inputs, and pretrained use, but does not train a capable model from scratch.
- Examples
- Demonstrates generated flowers, an astronaut riding an origami elephant, negative prompts, and varying diffusion-step counts.
- What you can do after
- Operate/configure a pretrained Stable Diffusion model to generate and steer images with prompts, negative prompts, and diffusion steps.
Read how Deep Learning with R, Third Edition explains text-to-image model →
The AI Pocket Book
A generative AI technique that produces images conditioned on textual descriptions.
- How it is used
- Generates images from natural-language descriptions, including practical artifacts such as logos.
- How deep it goes
- Explains conditional diffusion through inputs, supervised training, CNN reconstruction, and repeated denoising, without implementation details.
- Examples
- Uses Midjourney, DALL-E, captions, corrupted images, Gaussian noise, and logo creation as concrete examples.
- What you can do after
- Apply it via a text-to-image app to create an image or logo from a natural-language prompt.
Grokking AI Algorithms, Second Edition
Text-to-image generation is the process of synthesizing an image from a textual description.
- How it is used
- Uses paired images and captions to connect visual content with linguistic concepts for prompt-conditioned image synthesis.
- How deep it goes
- Explains dataset collection and caption quality, but stops short of model architecture or image-generation mechanics.
- Examples
- Uses a detailed red-tailed hawk caption, a vague bird label, synthetic imagery, and 3D-rendered scenes.
- What you can do after
- Evaluate captions and image sources for specificity, control, licensing, privacy, and copyright risks.
Read how Grokking AI Algorithms, Second Edition explains text-to-image model →
Related concepts
- diffusion modeluses · A diffusion model is a generative model that learns to produce data by reversing a process that progressively adds noise to data.
- sequence predictionkind of · Sequence prediction is the task of predicting or generating elements of an ordered sequence from preceding elements or contextual input.
- multimodalkind of · A multimodal model is a model that processes, generates, or relates information represented in two or more data modalities, such as text, images, audio, or video.
- Stable Diffusionincludes · Stable Diffusion is a family of latent diffusion models that generates and edits images from text and other conditioning inputs.
- conversational aikind of · Conversational AI is artificial intelligence that enables software systems to interact with users through natural-language dialogue.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.