Yongzhi Su

If you set this [arg](https://github.com/suyz526/ZebraPose/blob/06b53d2aab1f5e08ffc927b6c5a02e9913cb13f8/zebrapose/config/config_BOP_effnet/lmo/lmo_BOP_effnet_train.txt#L22) as True, the network will predict entire mask instead of visible mask.

Hi, thanks for your interest. You need to generate the ground truth for training. If your own dataset is organized in the same way as BOP data, then it can...

The binary code loss is only calculated for pixels within the predicted mask. For debugging, you can set `mask_binary_code_loss` to `False` to check if the binary code loss converges.

I don't have any logs available at this moment. Is your object symmetrical?