Dalle-3 via openai's website can directly annotate images by mid 2024
5
66
Ṁ478Ṁ130
Jul 1
4%
chance
1D
1W
1M
ALL
submit image, say of a cat
say something like "can you annotate this image with your notes on what parts of the cat you see? put 5 of them in bold black text"
dalle3 doesn't just "reproduce" an image using image => text => dalle3 generation again, (which looks very little like the original)
instead, it draws on top of the original image
so the original is basically present, but with modifications overlaid
it has to be through the normal UI
Interestingly, you can get it to output json blocks explaining what it sees in each region and then use python to look at that. But it's kind of messy and what it sees doesn't seem to be captured by that type of output well. I wonder if you can get the full embedding?
Today: fail
Get Ṁ200 play money
Related questions
Will OpenAI / Dall-E support real-time AI image generation in 2024?
70% chance
Will AI be able to generate an interactive web front-end by the end of 2024?
29% chance
OpenAI images have a useful and hard to remove watermark, by end 2024
30% chance
Dalle-3 via GPT-4 defaults to making more than one image at a time, by mid 2024
67% chance
Will OpenAI release a tool to identify images generated by DALL.E? (in 2024)
52% chance
Midjourney will be able to create images of people with different facial expressions by end 2024
48% chance
Leaks or press releases that AIs are doing end to end AI research by 2026
62% chance
Dall-e 4 comes out from OpenAI by mid 2024
23% chance
A public AI generator can correctly make images of complex 3d object stacking, Jun 1 2024
20% chance
OpenAI goes public or announces plans to by end 2024
38% chance