This is the training code used for the NViSII paper. You can also use this training script on the training data generated by the training script in scripts/nvisii_data_gen/

python -m torch.distributed.launch --nproc_per_node=1 train.py --network dope --epochs 2 --batchsize 10 --outf tmp/ --data ../nvisii_data_gen/output/output_example/

There is an accompanying dataset you can also use to train DOPE on the meat can with the shiny top. link here.

I also made an inference script that runs without any ROS components.

python inference.py

Look at the file for more information, similar to the ROS node everything is run through the yaml files in config_inference. It is very similar to the original code with some changes.

Check models.py as we are proposing different architectures.