About the network

[juipengchang]

Sorry, could you help me to solve some problems about the network in https://github.com/smallmax00/Graph_Region_Boudnary/blob/main/lib/model.py

1. The deep-level feature aggregation (8 x 8, 16 x 16), and the shallow and deep-level feature aggregation(8 x 8, 16 x 16, 32 x 32, 64 x 64) were both showed in the paper and used in the network. But only the shallow and deep-level feature aggregation was implemented in line 484-485. Is my understanding wrong?
2. The fusion of output seems only used low1 and low2 in line 209-216, 238-245. What is the reason? Thx!