Figure 8
From: A study on Shine-Muscat grape detection at maturity based on deep learning

Attention mechanism and five_conv structure. Note A 3 × 3 depth separable convolution is used in the five_conv module to reduce the number of parameters.
From: A study on Shine-Muscat grape detection at maturity based on deep learning
Attention mechanism and five_conv structure. Note A 3 × 3 depth separable convolution is used in the five_conv module to reduce the number of parameters.