Figure 3.

The architecture of recyclable trash sorting. (a) The recyclable trash sorting network, (b) the structure of the IPAM+ResNeXt block ($C_{\mathrm{in}}$, $C_{\mathrm{out}}$), where $C_{\mathrm{in}}$ and $C_{\mathrm{out}}$ represent the number of convolutional kernels, and GConv denotes group convolution with a group number of 32, (c) channel attention module (CAM) of IPAM, (d) the structure of IPAM, and (e) spatial attention module (SAM) of IPAM.