Vision models

AI systems engineered to interpret visual data (images, video), replicating human perception for automated analysis.

Vision models are deep learning architectures (CNNs, Vision Transformers) trained on massive datasets to enable machines to "see" and understand visual information. They power critical applications: object detection (YOLOv8), image classification (ResNet), and complex segmentation (SAM 2). These models extract meaningful features—like object identity and location—to automate tasks in autonomous vehicles, medical diagnostics, and real-time surveillance.

https://en.wikipedia.org/wiki/Computer_vision