This project is a collection of Python scripts to help work with datasets for deep learning. For example, visualizing annotation data associated with captured sensor images generated by NVIDIA Deep learning Dataset Synthesizer (NDDS) https://github.com/NVIDIA/Dataset_Synthesizer.
This module depends on OpenCV-python which currently doesn't work with Python 3.7; this module requires either Python 3.5 or 3.6.
Example of a dataset frame visualized using NVDU, showing axes and 3D cuboids for annotated objects.
pip install nvdu
Clone the repo
Using ssh path:
git clone ssh://[email protected]:12051/NVIDIA/Dataset_Utilities.git
Using https path:
git clone https://github.com/NVIDIA/Dataset_Utilities.git
Go inside the cloned repo's directory
cd Dataset_Utilities
Install locally
pip install -e .
nvdu_ycb command help download, extract and align the YCB 3d models (which are used in the FAT dataset: http://research.nvidia.com/publication/2018-06_Falling-Things).
usage: nvdu_ycb [-h] [-s] [-l] [ycb_object_name]
NVDU YCB models Support
positional arguments:
ycb_object_name Name of the YCB object to check.
optional arguments:
-h, --help show this help message and exit
-s, --setup Setup the YCB models for the FAT dataset
-l, --list List all the supported YCB objects
NOTE: If you don't run the nvdu_ycb --setup
before trying to use nvdu_viz, the visualizer will not be able to find the 3d models of the YCB object to overlay.
nvdu_viz command visualizes the annotated datasets using the NDDS format.
nvdu_viz [-h] [-a DATA_ANNOT_DIR] [-s SIZE SIZE]
[-o OBJECT_SETTINGS_PATH] [-c CAMERA_SETTINGS_PATH]
[-m MODEL_DIR] [-n [NAME_FILTERS [NAME_FILTERS ...]]]
[--fps FPS] [--auto_change] [-e EXPORT_DIR] [--auto_export]
[--ignore_fixed_transform]
[dataset_dir]
NVDU Data Visualiser
positional arguments:
dataset_dir Dataset directory. This is where all the images
(required) and annotation info (optional) are.
Defaults to the current directory.
optional arguments:
-h, --help show this help message and exit.
-a DATA_ANNOT_DIR, --data_annot_dir DATA_ANNOT_DIR
Directory path - where to find the annotation data.
Defaults to be the same directory as the dataset
directory.
-s SIZE SIZE, --size SIZE SIZE
Window's size: [width, height]. If not specified then
the window is sized to fit the resolution of the camera.
-o OBJECT_SETTINGS_PATH, --object_settings_path OBJECT_SETTINGS_PATH
Object settings file path.
-c CAMERA_SETTINGS_PATH, --camera_settings_path CAMERA_SETTINGS_PATH
Camera settings file path.
-n [NAME_FILTERS [NAME_FILTERS ...]], --name_filters [NAME_FILTERS [NAME_FILTERS ...]]
The name filter of each frame. e.g: *.png.
--fps FPS How fast to automatically change frame.
--auto_change When using this flag, the visualizer will automatically
change the frame.
-e EXPORT_DIR, --export_dir EXPORT_DIR
Directory path - where to store the visualized images.
If this is set, the script will automatically export
the visualized image to the export directory.
--auto_export When using this flag, the visualizer will automatically
export the visualized frame to an image file in the
`export_dir` directory.
--ignore_fixed_transform
When using this flag, the visualizer will not use the
fixed transform matrix for the 3d model.
NOTE: The nvdu_viz
script can work from any directory
- Visualize the current directory:
nvdu_viz
- Visualize a relative path:
nvdu_viz ../a_dataset
- Visualize an absolute path:
nvdu_viz ~/data/dataset
- Visualize different aspect of a frame using a filter:
nvdu_viz dataset_path --name_filters *.left.png *.right.png
- The camera and object settings files are in the image directory:
nvdu_viz image_directory_here -a annotation_directory_here
- The camera and object settings files are NOT in the image directory:
nvdu_viz image_directory_here -a annotation_directory_here -c camera_setting_path_here -o object_setting_path_here
F3 - Toggle the 2d cuboid
F4 - Toggle the 3d cuboid
F5 - Toggle the 3d models
F6 - Toggle the axes
F7 - Toggle the overlay frame name
1 - Render the 3d models normally
2 - Render the 3d models using only the edge lines
3 - Render the 3d models as point clouds
ESC - Quit the visualizer
Right - Go to the next frame
Left - Go to the previous frame
Up - Go to the next 100 frame
Down - Go to the previous 100 frame
Space - Toggle frame auto-changing
F12 - Toggle exporting the visualized frame to file