I was stunned, the first time I saw Dall-E 2. I really want to use it. While waiting for the beta invitation. I got my hand on other image generation, such as: StableDiffusion (via dream studio) and Midjourney.
This is not a comparing blog post between the three. It’s just my raw result from my experience.
- Easy UI. Very fun to play in discord.
- The image is not as accurate as I want though
StableDiffusion (via Dream Studio)
- Image is very far from what I want
- Found a local-installation (no credit applied?) https://github.com/divamgupta/diffusionbee-stable-diffusion-ui
So far the best in term of archiving what I want. I’m thinking to build my next project using the arts I generate from Dall-E 2
I think the skills we need to improve here is “how to write better prompt” and have a general knowledge in Artist’s word industry, to be a able to use the correct term.
Prompt Inspiration: https://arthub.ai/