Teach on Udemy

Turn what you know into an opportunity and reach millions around the world.

Learn More

Your cart is empty.

Keep shopping

NextLevel Images with Generative AI and Stable Diffusion GUI

Name: NextLevel Images with Generative AI and Stable Diffusion GUI
Rating: 4.4 (33 reviews)

Mastering Professional Image Creation: Explore real projects with Artificial Intelligence using a Graphical Interface

Created byJones Granatyr, Gabriel Alves, AI Expert Academy

Last updated 7/2024

English

What you'll learn

Generate complex images using only the Stable Diffusion graphical interface
Understand how to use Stable Diffusion parameters to obtain improved results
Create images using models provided by the Open Source community
Create book covers
Generate images with a higher level of realism
Create engaging characters
Add, edit, replace or delete elements in existing images
Transform simple images of sketches into professional generations
Create logos for companies, as well as stickers and prints for t-shirts
Learn creative use cases, generating illusions from logos or other types of images

Course content

9 sections • 85 lectures • 7h 29m total length

Course content4:38
Learn to generate professional images using open-source stable diffusion through a graphical interface, without coding, and explore case studies from book covers to logos, advanced editing, and image transformation.
Course materials0:04
What is Stable Diffusion7:09
Learn how stable diffusion, a latent diffusion deep learning model, generates high-quality images from text prompts using a graphical interface.
SageMaker Studio Lab and ngrok2:52
Preparing the graphical interface17:54
Open the Stable Diffusion graphical interface in SageMaker Studio Lab to generate images, configuring GPU runtime, running the setup script, and accessing the web UI via Ngrok.
Recommendations and common problems7:34
Maximize SageMaker Studio Lab usage by managing disk space and resolving common problems when using Stable Diffusion, including storage limits, temporary space, and symbolic links.
Generating the first images5:35
Perform the first stable diffusion tests in a graphical interface, using the sdxl base model and other models, with text prompts and adjustable width, height, and styles.
Parameters15:48
Explore how different scheduler algorithms influence image generation in stable diffusion, learn to use seeds for reproducible results, and adjust inference steps and cfg scale for prompt adherence.
Negative prompts4:28
Image resolution and proportions8:11
Learn how to select image resolution and aspect ratios for stable diffusion xl models, targeting 1024 by 1024 as default and avoiding 512 by 512 to prevent quality loss.
Prompt generator with ChatGPT6:45
Generating professional images10:34
Define what to generate and the purpose, choose the approach and model, craft prompts and prompt engineering, then apply post-processing, upscaling, inpainting, and final text.
Model biases11:59
Explore how training data shapes AI image generation and drives racial and gender biases, and learn prompting strategies to mitigate bias and promote diverse representations.

Introduction3:58
Explore book cover design with generative AI and stable diffusion, choosing portrait or landscape formats and proper dimensions, such as 832 by 1216 in a 13 by 19 proportion.
Initial prompt6:56
Keywords4:07
Style selection4:15
Use the style selector to quickly change image style and composition, centralizing elements with the keyword centralized at weight 1.5, while reviewing prompts, negative prompts, and color weights.
Variations4:08
Learn to create small variations of images with identical prompts by adjusting the original seed and variation seed, and by using variation strength and batch count for multiple outputs.
Upscaling and saving9:36
Generate a final book-cover image by refining prompts, negative prompts, and a digital art style, then upscale and save using image-to-image workflows, ultimate upscaler, and denoising strength 0.25.
Finishing the book cover5:16
Science fiction book 17:24
Learn to generate science fiction book covers with stable diffusion, using style prompts, community prompts, prompt books, and artist references for varied cover concepts.
Science fiction book 23:43
Generate three science fiction book covers by transforming descriptions into prompts with ChatGPT and a prompt generator. Space station Mars orbit and time travel through a vortex with vibrant colors.
Science fiction book 35:42
Learn to generate science fiction book covers using ChatGPT prompts, Dream Shaper Excel with Stable Diffusion GUI, and optimize steps, seeds, and prompts for high-quality results.
Epic fantasy book2:41
Non-fiction book2:30
Horizontal book covers2:59
Generate horizontal book covers in landscape mode by adjusting width and height to SD XL aspect ratios and resolutions, applying cyberpunk neon colors.
Creating scenarios1:34
Discover prompts to create fictionary, 3D, architectural, and realistic scenarios using stable diffusion, from living room layouts to a futuristic city and a Norway landscape.
Images with texts5:16
Learn to generate images with text using stable diffusion, download and load a text model, and combine prompts like text logo nebula with Laura and other extensions for better results.
Bonus: Python code4:50
Explore generating images with stable diffusion in Python via Google Colab notebook, install diffusers and transformers, load models, select schedulers, and save and upscale outputs.

First images5:35
Photographs5:13
Realistic models 17:33
Explore realism-focused diffusion models, including sd xl and 1.5 variants, and master prompts, negative prompts, and model-specific dimensions and schedulers in the Stable Diffusion GUI.
Realistic models 24:08
Generate images with realistic models using stable diffusion 1.5 at 512 by 512, apply a negative prompt to reduce blur, and compare results across Realistic Vision and Epic Realism.
Improving faces, hands, and eyes6:52
Learn to improve face and hand quality using face and hands yolo detection, bounding boxes, and targeted model selection; combine text inversions and embeddings for sharper results.
Photographs styles4:43
Poses and actions4:00
Explore prompts to describe age, hair, clothes, facial expressions, accessories, and moods, while guiding eye contact and poses for realistic image generation with stable diffusion.
Photo framing2:33
Background and environment1:19
Lighting3:05
Camera angles0:58
Explore camera angles in generative imaging, including level shots aligned with the eyes, low angles from below, high angles from above, and the Dutch angle for a diagonal, distorted perspective.
Camera settings6:23
Explore camera settings and the shot on technique to simulate cinema and retro looks using diverse cameras, film types, and focal lengths.
Filters and effects1:49
Explore filters and effects, including black and white, bokeh, and lens flare. Use color tweaks like purple lights, blue lights, and red lights, and apply long exposure for dynamic images.
Complete prompts2:49
Craft complete prompts by combining style, subject details, lighting, and camera angle. Use filters, effects, and camera type with a default negative prompt and default parameters.
Upscale for realistic photos4:02
Learn to upscale realistic photos by choosing the right algorithms and parameters, doubling image size from 512x512 to 1024x1024 with controlled denoising, while focusing prompts on the environment.
Poses with controlnet8:01
Learn to control stable diffusion using control net and Open Pose key-point extraction to generate pose-consistent images from sketches, with practical steps for selecting control types, prompts, and settings.

Astronaut panda - introduction6:24
Explore character creation with stable diffusion and generative AI, using Cosmo the Adventures Panda astronaut as a hands-on example across literature, games, cinema, education, and marketing in NextLevel images course.
Astronaut panda - first tests3:37
Astronaut panda - styles1:59
Astronaut panda - action and location2:43
Highlight the action and location of a panda in a blue astronaut suit at a spacey coffee shop. Use cinematic prompts and test 832 by 1216 dimensions with ChatGPT.
Astronaut panda - more styles4:42
The sorceress5:16
The wise2:35
Create options for a wise gray-haired wizard named Gray Leaf from the ancient forests. Use ChatGPT prompts and stable diffusion to test variations and adjust prompts for different visual results.
Other models 15:30
Explore prompt comprehension across models, including sd xl and open dall-e, and learn to choose stable diffusion options for game characters, realistic and fantasy images.
Other models 26:54
Explore testing custom modules with stable diffusion GUI, using the gallery to copy keywords and prompts, and fine-tune results with textual inversion and parameters like guidance scale and steps.
Consistent faces with Reactor10:17
Image variations4:25
Generate image variations with image-to-image prompts and negative prompts, adjusting denoising strength and seeds to control face consistency. Adjust batch size and environments to produce multiple outputs.

Inpainting - introduction5:56
Learn to edit images with AI using inpainting to add, modify, or delete elements and outpainting to extend regions. Use denoising strength and ControlNet modules to control edits and reconstruction.
Inpainting in practice15:43
Add, modify, and delete objects8:22
Add, modify, and delete objects using stable diffusion GUI and inpaint. Explore image-to-image workflows, prompts, and parameter options to influence outcomes.
Editing realistic images6:03
Outpainting5:39
IP Adapter8:11
Use IP adapter to transfer features from an input image to a target image, enabling face changes in stable diffusion with inpaint and control net.

Inpaint sketch6:08
Learn to use the inpaint sketch tool in the img to img tab with Stable Diffusion and the Shaper XL model, drawing apples on a wooden table to generate results.
Photopea extension5:12
Scenario generation5:03
Scribble and Lineart - intuition3:01
Explore two algorithms for turning sketches into complex images: scribe and line art, using stable diffusion in a gui. Learn inputs for each model and how to enable both tools.
Sketches into complex images5:19
Architecture images5:54
Sketches into characters2:55
Sketches into products3:40
Images with scribbles and lines5:19
Increase realism2:57
Edge detection5:09
Learn to extract image edges to feed control net for image generation. Experiment with Open Pose and canny edge detection, adjusting thresholds and control weight for different results.

Logo for coffee shop 15:30
Create vector logos and logo marks for brands using stable diffusion, with a Mars Coffee case study, exploring minimalist vector designs, prompts, color controls, backgrounds, and logo variations.
Logo for coffee shop 26:22
Other logos8:49
Generate and refine logos using stable diffusion GUI by adjusting sampling steps, background colors, and prompts, and compare results from logo-specific models like Logo Redmond v2.
Logo post-processing4:15
Learn logo post-processing workflow: upsample to 2048 with low denoising, apply Ultimate SD upscale and Resr upscaler, then remove backgrounds and create a design in Canva for Mars Coffee.
Stickers1:45
Arts for t-shirts4:54

Requirements

No prerequisites, as the graphical interface is used to generate the images. A ready-to-use code is available to start the interface

Description

This course dives deep into the fascinating world of generative artificial intelligence, with a special focus on the powerful technique known as Stable Diffusion using a friendly Graphical User Interface (GUI). Learn how to master this revolutionary technology to create stunning images, from book covers to charming characters and much more. With a practical and results-oriented approach, you will be guided through a learning journey that covers everything from basic concepts to advanced image manipulation techniques. Check out what you will learn below:

1. Fundamentals of Generative AI and Stable Diffusion

- Comprehensive introduction to Stable Diffusion and its application to AI-generated images.

- Exploration of SageMaker Studio Lab and ngrok for efficient configuration of the development environment.

- Interface preparation and troubleshooting common errors for a better workflow.

- Using parameters and negative prompts to refine image generation.

- Understanding the bias of models and their influence on content generation.

2. Cover Creation

- Step-by-step guide to creating eye-catching book covers using generative AI.

- Selection and manipulation of styles through initial prompts and keywords.

- Exploration of variations and use of upscale to improve print quality.

- Use of post-processing techniques, such as adding texts to the final art.

- Customization of covers for different literary genres, from science fiction to epic fantasy.

- Generation of images with readable texts and creative and diverse scenarios.

- Creation of other types of covers in different sizes, which can be used, for example, for video thumbnails or article covers.

3. Realistic images generation

- Advanced techniques to generate realistic images.

- Enhancement of details such as faces, hands and eyes for a more authentic look.

- Exploring photography styles, poses, angles, framing and lighting to create visually captivating compositions.

- Use of comprehensive prompts and upscale to obtain high-quality realistic images.

- Customization of poses with ControlNet for greater control over the final image.

4. Character Creation:

- Creation of unique characters, from astronaut pandas to sorceresses and wizards.

- Experiment with different styles and actions to bring the characters to life.

- Use of ReActor to ensure consistent faces and image variations.

- Exploration of advanced editing techniques for refinement and customization.

5. Advanced AI Image Editing

- Introduction to inpainting and advanced editing techniques.

- Adding, modifying and deleting any object in the image to achieve precise manipulation.

- Exploration of outpainting and IP Adapter to expand creative possibilities and make complex structural changes in a practical way.

6. Transforming simple images into complex ones

- Techniques to bring simple sketches and illustrations to life, transforming them into complex and refined art in an instant.

- Using an image-to-image approach to increase the realism and the level of detail in any existing image.

- Exploring the use of ControlNet models to transform the style of the image.

7. Creation of Logos and other vector arts

- Creation of professional logos for any type of company, entity or brand.

- Improving the logo using post-processing techniques, leaving it ready for professional use.

- Exploration of techniques to create stickers, t-shirts prints and other forms of promotional material.

8. Creative Use of Images

- Technique to gain full control over image composition.

- Creative application of images generated in various areas, including illusions using logos, geometric shapes, texts and other forms of artistic expression.

Upon completing this course, you will be equipped with the skills needed to utilize Generative AI and Stable Diffusion in a variety of creative contexts, from book cover design to character creation and custom illustrations. A totally practical course using a web User Interface to generate images!

Who this course is for:

Graphic designers: Professionals who want to expand their design skills using advanced generative artificial intelligence techniques to create unique and engaging images.
Visual artists: People interested in exploring new forms of artistic expression and expanding their creative repertoire through the use of emerging technologies.
Game developers: Individuals involved in creating electronic games who seek to integrate original and captivating visual elements into their projects.
Marketing and advertising professionals: Marketing specialists looking to explore new ways to create attractive and engaging visual content for advertising and branding campaigns.
Artificial Intelligence enthusiasts: Individuals interested in the intersection of art and technology, who wish to explore the creative possibilities offered by generative AI and Stable Diffusion.

NextLevel Images with Generative AI and Stable Diffusion GUI

What you'll learn

Explore related topics

Course content

Introduction13 lectures • 1hr 44min

Book covers16 lectures • 1hr 15min

Realistic images16 lectures • 1hr 9min

Character creation11 lectures • 54min

Image editing6 lectures • 50min

Simple to complex images11 lectures • 51min

Vector arts and logos6 lectures • 32min

Creative use4 lectures • 13min

Final remarks2 lectures • 3min

Requirements

Description

Who this course is for: