Skip to main content
. 2025 Oct 7;16:1677883. doi: 10.3389/fpls.2025.1677883

Figure 9.

Flowchart illustrating a neural network architecture divided into three sections: backbone, neck, and head. The backbone section processes input images through a series of convolutional layers, labeled with dimensions. The neck section includes concatenation and upsampling processes, transferring information to the head section. The head contains detection layers, producing outputs with labeled dimensions. Arrows indicate data flow between processes.

Structure of YOLO_P2.