Inside the modelnet_list directory, run modelnet_list_generator.py. modelnet_list_generator.py
expects different args which are as
- modelnet_data_path
- train_num
- test_num
- val_num
Running the above file will generate train, test and validation files consisting of CAD models from modelnet dataset.
To generate depth maps from these models run generate_h5_data.sh script which take three args,
output path to save h5 depth data, number of views for each model, output path to save raw depth data.
e.g., sh generate_h5_data.sh /home/rash8327/Desktop/modelnet_off_h5 100 /home/rash8327/Desktop/modelnet_off_depth