Computer Vision
A field of AI that enables computers to interpret and understand visual information from images and videos.
Giving Machines the Ability to See
Computer vision teaches machines to extract meaning from pixels. When you look at a photo, you instantly recognize faces, objects, text, emotions, and spatial relationships. Computer vision systems aim to replicate this capability - and in some narrow tasks, they now surpass human accuracy.
The field started with simple tasks like edge detection and evolved into sophisticated systems that can describe what's happening in a photo, track objects through video, or generate realistic images from scratch. Deep learning, particularly convolutional neural networks, drove most of these advances.
Real-World Applications
Computer vision is already embedded in daily life. Your phone unlocks with your face. Cars detect pedestrians and read speed limit signs. Medical systems spot tumors in X-rays. Manufacturing lines catch defects. Social media tags your friends automatically.
The technology also raises important questions. Facial recognition enables mass surveillance. Deepfakes spread misinformation. Bias in training data leads to systems that work poorly for certain demographics. As computer vision becomes more powerful, these concerns become more urgent. The technical capability to see isn't enough - we also need thoughtful policies about how that capability gets used.