Using Craiyon and DeepAI to test text-to-art AI prompts | Why Try AI

Why Try AI?
22 Sept 202214:03

TLDRIn this episode of 'Why Try AI', the host explores free AI art generators, Craiyon and DeepAI, to test and refine text prompts for creating art without the need for registration. The video demonstrates how to iteratively improve prompts to achieve a desired dystopian image of a giant mouse over a city, emphasizing the importance of consistency over perfection in these tools. The host also advises on taking notes of the prompts and outcomes for future reference, concluding that while the results may not be as polished, they serve as a cost-effective way to experiment and hone skills before using more advanced models.

Takeaways

  • 🎨 There are many AI art generators that can create impressive art, but they often require sign-up or payment for extensive use.
  • 🆓 Craiyon and DeepAI are two free tools that allow for text-to-art generation without the need for registration.
  • 📝 Both tools focus on the text prompt as the core of communication with AI art generators, which is an important skill to master.
  • 🔄 They provide multiple images for each text prompt, which is useful for testing consistency and iterating on ideas.
  • 🚫 The quality of the images from these free tools is not as high as more sophisticated models, so they are not suitable for benchmarking the capabilities of AI art.
  • 🔍 Consistency in the output is more important than perfection when using these tools to refine text prompts.
  • 📸 Craiyon has a feature to take a screenshot of all generated images along with the text prompt, aiding in the review process.
  • 🔄 The process involves starting with minimal information and iteratively adjusting the text prompt based on the AI's output.
  • 🖌️ Adding specific adjectives and descriptors can help guide the AI to create images closer to the desired outcome.
  • 👍 Once a satisfactory text prompt is found, it can be used in more advanced programs for higher quality results.
  • 💡 The video demonstrates how to work with these tools effectively, emphasizing the importance of iteration and experimentation.

Q & A

  • What is the main purpose of the video episode of 'Why Try AI'?

    -The main purpose of the video episode is to demonstrate how to use free AI art generators to test out ideas and text prompts, and to improve them for use in more sophisticated models.

  • Why are some AI art generators not accessible without a sign-up process or payment?

    -Some AI art generators require a sign-up process, registration, or payment because they offer more advanced features and capabilities, which come at a cost to maintain and operate.

  • What are the two free tools introduced in the video for generating AI art without signing up?

    -The two free tools introduced are Craiyon (formerly known as Dali mini) and Deep AI text to image generator.

  • How do Craiyon and Deep AI differ from other AI art generators in terms of user interaction?

    -Craiyon and Deep AI differ by focusing solely on the text prompt, eliminating the need for users to adjust settings or choose algorithms, making the process simpler and more accessible.

  • Why is it important to test text prompts in AI art generators?

    -Testing text prompts is important because it helps users refine their prompts to achieve the desired outcome, which is a crucial skill when working with AI art generators.

  • What is the advantage of getting multiple images for each text prompt in Craiyon and Deep AI?

    -The advantage is that it allows users to see if their text prompt is consistent in producing the desired result, which is helpful for refining the prompt before using more advanced models.

  • What are the two caveats mentioned about using Craiyon and Deep AI for AI art generation?

    -The two caveats are that the quality of the images produced may not be as polished as those from more advanced models, and the results may not be consistent across different AI art generators due to differences in algorithms.

  • How does the video demonstrate the process of refining a text prompt for AI art generation?

    -The video demonstrates by showing an iterative process where the user starts with a basic prompt and gradually adds or modifies terms to achieve a more specific and desired outcome.

  • What is the significance of noting down the text prompts and the resulting images during the AI art generation process?

    -Noting down the text prompts and images is significant for tracking progress, understanding the impact of different terms on the outcome, and refining the prompts for better results in the future.

  • How does the video suggest using adjectives and style modifiers to improve AI-generated art?

    -The video suggests using adjectives like 'sinister' and 'intricate' to convey the desired mood and level of detail, and style modifiers like 'by Frank Miller' to mimic the aesthetic of specific artists.

  • What is the final recommendation made in the video regarding the use of free AI art generators?

    -The final recommendation is to use the free AI art generators like Craiyon and Deep AI to quickly iterate and refine text prompts before moving on to more sophisticated programs that offer more control and higher-quality results.

Outlines

00:00

🎨 Introduction to Free AI Art Generators

The video begins with an introduction to the concept of AI art generators, highlighting their ability to create stunning artwork that rivals human artists. The host encourages viewers to explore these tools, despite the common requirement of registration or payment to access more features. Two free AI art generators, Crayon (formerly Dali mini) and Deep AI Text to Image Generator, are introduced as alternatives that allow users to generate images without any sign-up process. These tools focus on text prompts, simplifying the user experience by eliminating complex settings, and provide multiple image outputs per prompt, aiding in the iterative process of refining prompts for better results.

05:02

🖌️ Iterative Process of AI Art Generation

The host demonstrates the iterative process of refining text prompts to achieve desired AI-generated images. Starting with a simple prompt of a 'giant mouse looking down on a city movie poster,' the video shows how the initial output was far from the intended dark, dystopian theme. Through a series of adjustments, such as adding 'photo' and 'from the ground,' the host aims to guide the AI towards a more realistic and sinister image. The process involves trial and error, with the host noting the importance of documenting each prompt and its corresponding images for future reference. The goal is to find consistency in the AI's interpretation of the prompts.

10:04

📈 Refining AI Art Prompts for Better Results

Continuing the iterative process, the host experiments with various text prompts to achieve a more sinister and realistic image of a giant mouse over a city. The use of terms like 'silhouette,' 'scary,' and 'looming' helps guide the AI towards the desired aesthetic. The host also explores the addition of specific artistic styles, such as 'by Frank Miller,' to see if it enhances the image. However, not all additions are successful, and the host demonstrates the importance of removing elements that do not contribute positively to the final result. The video concludes with a comparison of the refined prompt's results between Crayon and Deep AI, showing the consistency in the AI's interpretation and the potential for further refinement in more sophisticated programs.

Mindmap

Keywords

💡AI art generators

AI art generators refer to software or platforms that use artificial intelligence to create visual art based on text prompts or other inputs. In the video, the host discusses using these tools to quickly test and refine text prompts to produce desired images, highlighting their potential to evolve and improve with time. The script mentions that while many AI art generators require registration or payment, the two tools featured, Craiyon and DeepAI, offer free and accessible alternatives for users to experiment with.

💡Text prompts

Text prompts are the textual descriptions or instructions given to AI art generators to guide the creation of an image. The video emphasizes the importance of mastering the art of crafting effective text prompts as the core interaction with AI art generators. The host demonstrates the iterative process of refining text prompts to achieve a desired outcome, such as a 'dark dystopian movie poster of a giant mouse hovering over a city'.

💡Craiyon

Craiyon, formerly known as Dali mini, is one of the free AI text-to-image generators mentioned in the video. It allows users to input text prompts without the need for registration and generates nine images per prompt. The script uses Craiyon to illustrate the process of testing and improving text prompts to achieve more consistent and desired AI-generated images.

💡DeepAI

DeepAI is another free AI text-to-image generator highlighted in the video script. It operates similarly to Craiyon by accepting text prompts and generating four images per prompt. The video uses DeepAI as an additional tool for testing text prompts, emphasizing the value of comparing results across different AI platforms to refine the prompts further.

💡Image consistency

Image consistency refers to the reliability with which an AI art generator produces similar outcomes from the same text prompt. The video script discusses the importance of achieving consistency as a sign that a text prompt is effective. It suggests that if multiple images from a single prompt align with the desired theme or concept, the prompt is likely well-crafted.

💡Dystopian

Dystopian is a term used to describe a society characterized by oppression or control, often depicted in literature and art as a dark and grim scenario. In the context of the video, the host aims to create a dystopian-themed image of a giant mouse over a city, using this concept to guide the refinement of the text prompts for the AI art generator.

💡Sinister

Sinister is an adjective used to describe something with a threatening or evil quality. The video script includes 'sinister' as part of the desired aesthetic for the AI-generated image, aiming for a dark and ominous atmosphere. The host uses this term in the text prompts to guide the AI towards creating an image that conveys a sense of threat or unease.

💡Photorealism

Photorealism is a style of art that strives to achieve the same high level of detail and resemblance to real objects as a photograph. In the script, the host initially seeks a photorealistic image of a giant mouse looking down at a city. The term is used to communicate the desired level of detail and realism in the AI-generated output.

💡Silhouette

A silhouette refers to the dark shape and outline of an object or person cast against a lighter background, often used in art to create a dramatic effect. The video script mentions using 'silhouette' in text prompts to guide the AI towards creating an image with a dark, ominous figure against a lighter backdrop, enhancing the sinister and dystopian theme.

💡Iterative process

The iterative process is a series of repeated steps or trials with the goal of achieving an improved outcome. In the context of the video, the host demonstrates an iterative process of testing and refining text prompts with AI art generators to gradually get closer to the desired image. This process involves making adjustments based on the results of each iteration.

💡Frank Miller

Frank Miller is a renowned comic book artist and writer, known for his work on 'Sin City' and 'Batman: The Dark Knight Returns,' which often features dark, black-and-white visuals. In the script, the host experiments with adding 'by Frank Miller' to the text prompt to see if the AI can mimic his distinctive style. However, the results are inconsistent, leading to the decision to remove this modifier from the prompt.

Highlights

Introduction to using AI art generators for testing text prompts without the need for sign-up or payment.

AI art generators can create art that rivals the work of real artists, but often require sign-up processes or payment.

Two free tools, Craiyon and DeepAI, allow for quick testing of text prompts without registration.

Craiyon and DeepAI provide multiple images per text prompt, aiding in consistency checks.

The importance of mastering text prompts when using AI art generators.

Caveat: Free AI models like Craiyon and DeepAI may produce lower quality images compared to paid models.

Consistency in AI-generated images is more important than perfection when testing prompts.

The process of iteratively refining text prompts to achieve the desired AI-generated image.

Using the 'screenshot' feature in Craiyon to document the evolution of text prompts and results.

Experimenting with different terms to achieve a more realistic and sinister look in AI-generated images.

The challenge of using relational terms in text prompts and the impact on AI-generated image outcomes.

Adding adjectives like 'scary' and 'sinister' to refine the tone of the AI-generated image.

The trial and error process of adding and removing terms to find the best text prompt combination.

Using the term 'intricate' to encourage the AI to add more details to the generated image.

Testing the impact of specifying an artistic style, such as 'by Frank Miller', on AI-generated images.

The importance of removing elements that do not contribute positively to the AI-generated image.

Comparing the results from Craiyon and DeepAI to assess consistency across different AI models.

Finalizing a text prompt that is close to the desired outcome and ready for use in more sophisticated AI models.

The efficiency of using free AI tools for quick iterations and refinement of text prompts before moving to paid models.