Help choosing ai image generation model
Hello guys. I have seen a couple project here built with wasp that involves image genaration. Never done it before and by doing some research i have found lots of models. So i am trying to narrow down my options.
My use case:
I want to generate images that illustrate scenarios related to to traffic rules, road signs, priority... school driving stuffs. And from that image user will have to answer whats happening in that feature, like which car should go first? The images can be 2D, since its whats present in most of the books, and quality just enough for one to see details like the sign, the road... Any siggestions?
@Vinny (@Wasp) saw that you were doing a project that relates with image generation, can you please help me out here???
4 Replies
Hi @Stefan Vitória. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
----------------------
I can see you're looking to implement AI image generation for traffic scenarios in a Wasp app. While the knowledge sources provided don't contain specific information about which AI image generation model would be best for your traffic rule scenarios, I can share what we know about using AI with Wasp.
Wasp is primarily a framework for building full-stack web applications with React, Node.js, and Prisma. For AI image generation, you would typically integrate with external AI services rather than implementing the image generation directly in Wasp.
From the knowledge sources, we can see that Wasp supports:
- Building multi-modal applications that leverage computer vision and image capabilities Discord thread
- Integration with AI services through APIs
- RAG (Retrieval Augmented Generation) implementations, as shown in their example app: ask-the-documents
For your specific use case of generating traffic scenarios, you would need to:
1. Choose an external image generation API (like OpenAI's DALL-E, Stable Diffusion, or Midjourney)
2. Integrate it with your Wasp app through API calls
3. Provide specific prompts to generate the traffic scenarios you need
Unfortunately, the knowledge sources don't provide specific recommendations for which model would work best for traffic rule illustrations. You might want to check the Wasp documentation or ask in the Wasp Discord community for more specific recommendations from users who have implemented similar features.
If you're new to Wasp, you might want to start with their tutorial or templates to get familiar with the framework first: Wasp Tutorial
kapa.ai
----------------------
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
----------------------
Sorry @Stefan Vitória, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist:
Perhaps try again in 15mins?
@Stefan Vitória I used Replicate and Flux Lora Dev. Loras are great for getting a consistent style.
you'd probably want to use a lora that has a simple, line illustrated style
@Vinny (@Wasp) thanks. I will take a look. In you case you had to "train" it to feet your specific needs?
no I could use LoRAs that others trained/created. check out hugging face for a bunch of em. or even on replicate there should be some.