郑天宇

Do you recently have any dev plans for Bayer format image like RGGB GBRG?

19

: )

not-edge loss value is too small, edge loss is nan/inf

4

Hi, I tried your affinity loss (not adaptive) as my loss function, my network is DeeplabV3+, MobileNet, My own dataset. I set margin=3.0, lambda1=1.0, lambda2=1.0 But there is something wrong...

Where is the last layer (1280 channels) in MobileNetV2 implementation?

1

In the model.py, I didn't find the last layer (1280 channels) after No.16 inverted_res_block. Why? It exists in MobileNetV2 paper and tensorflow.deepLab implementation.