model

[hexingwei]

The model you introduced in your published paper seems different from the model in your code? Can you explain it more concretely? Only the Aggregation Layer (CNN layer), which you described in your paper, has trainable parameters. Thanks very much.