AI Unveiling Multi-Attacks in Image Classification: How One Adversarial Perturbation Can Mislead Hundreds of Images