Markus Schön

No, I did not get an updated version yet.

> Hi @soldierofhell, thanks for your interest in our work. Yes, we are still preparing the instance/panoptic repo. Please expect a release in mid- to late January 2021. Any news...

1. The models provided in the model zoo were trained using the commands in the [Reproduce Results](https://github.com/markusschoen/MGNet/blob/main/GETTING_STARTED.md#reproduce-results) section. To generate pseudo labels for the KITTI dataset, you can either use...

A negative total loss is normal, it is caused by the uncertainty weighting. Can you post an example of a fuzzy depth map obtained by your model? (You can use...

Your results are strange, can you please double check your setup? When I run inference on this particular image using the command `python tools/demo.py --config-file configs/MGNet-KITTI-Eigen-Zhou.yaml --input "0000000069.png" --output ~/Desktop...

To train on your own dataset, you can follow the instructions for custom datasets in [detectron2](https://detectron2.readthedocs.io/en/latest/tutorials/datasets.html). Yes, you wont need pseudo labels, if you only train the depth task.