emotiCAPTCHA

[0x5844]

Hi lovely people!

I came around Dia a while back; on the first try I got hooked into the idea of using it for captcha purposes. I made a small project; it generates a short audio dialogue using Dia. This dialogue includes a non-verbal emotional cue (like a laugh, sigh, etc.). To pass the CAPTCHA, the user needs to:

1. Identify which speaker performed the non-verbal cue.

2. Select the specific non-verbal cue that was performed (using an emoji-based selection).

3. Determine the general topic of the conversation (testing basic comprehension).

The idea is to achieve a potentially more robust CAPTCHA by relying on the understanding of nuanced audio and emotional context. All the nootbokes and code could be find at:

• GitHub Repository: https://github.com/0x5844/emotiCAPTCHA

• Colab Notebook: https://colab.research.google.com/drive/1vUIYs7dVdndkadxgajEevIvWIKm8Rg3T

I'd love to hear any thoughts, feedback, or ideas you might have. Happy coding ;D