1. This article belongs to the YOLOV5/YOLOV7/YOLOV8 improvement column and contains a large number of improvement methods. The improvement methods are mainly proposed based on the latest articles in 2023 and articles in 2022.
2. Provide more detailed improvement methods, such as adding attention mechanisms to different positions of the network, which facilitates experiments and can also be used as an innovation point in the paper.
2. Point increase effect: Change to MPDIOU to achieve effective point increase!
Table of contents
Bounding box regression (BBR) is widely used in target detection and instance segmentation and is an important step in target localization. However, most existing bounding box regression loss functions cannot be optimized when the predicted box has the same aspect ratio as the groundtruth box, but completely different width and height values. In order to solve the above problems, we fully exploit the geometric characteristics of horizontal rectangles and propose a new bounding box similarity comparison metric MPDIoU based on the minimum point distance. This metric includes all relevant factors considered in existing loss functions, namely overlap. or non-overlapping area, center point distance, width and height deviation, while simplifying the calculation process. On this basis, we proposed a bounding box regression loss function based on MPDIoU