Zero-Shot & Open-Vocabulary Detection

Zero-shot and open-vocabulary detection find objects from a text prompt without training on those classes. These posts cover models like YOLO-World and Grounding DINO and when they are accurate enough to use.