PerilPix: A Curated Dataset for Detecting Extreme and Normal Selfies

The dataset contains 2,000 images equally divided into "Normal Selfie" and "Extreme Selfie" categories, with 1,000 images each. "Normal Selfie" images were sourced from UCF's public dataset, while "Extreme Selfie" images were collected from social media and online platforms. Manual validation ensured accurate classification for reliable model training.