Non-Maximum Suppression

Non-Maximum Suppression (NMS) is a technique used in object detection to refine results. When a model identifies an object, it may produce multiple overlapping boxes around the same item. NMS helps by selecting the most confident detection (highest score) and suppressing—or discarding—other overlapping boxes that likely refer to the same object. This process ensures that each object in an image is represented by a single, clear bounding box, making the detection results more accurate and easier to interpret.