Computer Vision

Computer Vision (CV) is the branch of AI that gives machines "eyes." Recognizing a cat in a photo, letting a self-driving car spot a pedestrian, or having a factory camera catch a defective product — any AI technology that works with visual information falls under this umbrella.

Along with NLP (natural language processing), which handles human language, it forms one of the two main pillars of AI. Many things we already use daily belong here too, from face-recognition unlock on smartphones to image search on search engines to reading medical scans.

Recently, as multimodal AI that handles text and images together has become mainstream, computer vision and language models are merging. The result is being able to paste a photo into a chatbot and ask, "What is this?" When news mentions a "vision model" or "visual AI," it's referring to computer vision technology.

"A quality inspection system based on computer vision cut the defect rate in half" — this means cameras and AI, instead of human eyes, are automatically catching defective products.

1. Paste any photo into a chatbot and ask, "Describe what's in this picture."
2. The AI will read out objects, places, even the mood — this is "image recognition," the basic task of computer vision.
3. Go a step further: take a photo of a handwritten note and ask, "Convert this note to text." Character recognition (OCR) is also part of computer vision's job.