image

Definition

Two-dimensional pixel data, the core modality of computer vision. It anchors both recognition tasks — classification, detection, segmentation — and generative ones such as text-to-image synthesis. The contrast with video is not cosmetic: adding a time axis multiplies compute, introduces temporal consistency requirements, and changes which architectures remain practical.

Explainers using this term

• Paper Walkthrough: LLaDA-Image — Building the Visual Prior from Images Alone, Then Distilling to 2–4 StepsLLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes
• Document AI and OCR Today — How an LLM Ends Up Reading Your InvoicesLayoutLM: Pre-training of Text and Layout for Document Image Understanding
• Neural Compression — The Codec That LearnsEnd-to-end Optimized Image Compression