Categories: Easy – ML:20% CTF:80% #mllm
A friend of mine sent me a link to a service that determines the make and model of a car, as well as its ‘coolness’ rating.
His Ferrari 250 GTO got 100 points out of 100.
Let’s see if I can beat his score.
A rating of 1337 seems like a perfect fit.
In this challenge, there is an image processing mechanism that works in the background and scores the cars loaded into the system. The description asks us to beat this process and make the car’s score 1337. We need to experiment to understand how the mechanism scores.
Okay, let’s try this with a picture of a car, but let’s do it a little differently, for example let’s change the font and the samples types.
Let’s try it without a picture of the car.
It also seems from here that it may be enough to determine this method of deception only by writing.
Let’s see what comes back to us if we define objects correctly.
It didn’t work and I searched right prompt for this. Although it took a while to find the exact desired prompt but, I finally managed to trick it with the right prompt.
Flag: aictf{b35T_Car_I5_Y0Ur_F1rS7_Car}
