How we used GPT-4o for image detection with 350 similar illustrations

#1

Post by olup » Fri, Jan 10, 2025, 9:02 PM UTC

How we used GPT-4o for image detection with 350 similar illustrations
olup-blog.pages.dev

Re: How we used GPT-4o for image detection with 350 similar illustrations

#2

Post by olup » Fri, Jan 10, 2025, 9:02 PM UTC

First time for me posting this kind of story - I thought it would make an interesting case on solving a hard computer vision problem with a crafty product engineer team.

Re: How we used GPT-4o for image detection with 350 similar illustrations

#3

Post by caioariede » Mon, Jan 13, 2025, 9:50 PM UTC

olup wrote: post #2

First time for me posting this kind of story - I thought it would make an interesting case on solving a hard computer vision problem with a crafty product engineer team.

Just a small feedback… I have switched to the reader mode because the font used is very challenging to read for me.

Re: How we used GPT-4o for image detection with 350 similar illustrations

#4

Post by vessenes » Mon, Jan 13, 2025, 9:51 PM UTC

Thanks for the “bitter lesson” news from the frontlines. Curious; did you experiment with 4o as the sole pipeline? And of course as I think you mention, it would be interesting to know if say llama 8b could do a similar job as well.

Congrats on shipping.

Re: How we used GPT-4o for image detection with 350 similar illustrations

#5

Post by Melomomololo » Mon, Jan 13, 2025, 9:56 PM UTC

[dead]

Re: How we used GPT-4o for image detection with 350 similar illustrations

#6

Post by schappim » Mon, Jan 13, 2025, 10:01 PM UTC

I would love to see the prompt / image data sent to GPT-4o!

Re: How we used GPT-4o for image detection with 350 similar illustrations

#7

Post by gazchop » Mon, Jan 13, 2025, 10:12 PM UTC

I hear a lot of qualitative speak but nothing quantitative.

Re: How we used GPT-4o for image detection with 350 similar illustrations

#8

Post by GaggiX » Mon, Jan 13, 2025, 10:13 PM UTC

Is there a reason to choose VGG16 over more modern models?

Re: How we used GPT-4o for image detection with 350 similar illustrations

#9

Post by saint_yossarian » Mon, Jan 13, 2025, 10:16 PM UTC

I mean, cool tech, but why not just print a QR code next to each illustration?

Re: How we used GPT-4o for image detection with 350 similar illustrations

#10

Post by kredd » Mon, Jan 13, 2025, 10:30 PM UTC

A bit tangential, but I think we will see a good chunk of small teams building competing products in different software business segments, by just doubling on productivity and offering a cheaper option due to less operational overhead (reads: paying engineers). I can think of at least two businesses that can be competed in costs if the team can automate a good chunk of it.