phi-lab-rice/GRADE_Dataset

GRADE dataset

This dataset release contains raw recordings for GRADE: Single-Frame Generative Radar Depth Estimation Under Visual Degradation, accepted to ACM MobiCom 2026. See the paper and project website.

The recordings were made with a TI IWR1843BOOST radar (with DCA1000EVM), a ZED 2i stereo camera, and a DJI Action 5 Pro camera. Evaluation sequences also used a MAX30105 smoke sensor. This release currently contains only non-video data. We are anonymizing the DJI and ZED video recordings and will release them soon.

Folders and splits

GRADE_Train_Raw/       50 sequence folders
GRADE_Eval_Raw/        20 smoke evaluation sequence folders
processing_code/       processor.py, requirements.txt, and utils/

Split rule: Within GRADE_Train_Raw, the 10 folders with Dell or Keck in their names are for validation, not model training. This includes the Smoke-Dell-* and Smoke-keck-* folders. The other 40 folders are for training. All 20 folders in GRADE_Eval_Raw are for evaluation. Do not treat every folder under GRADE_Train_Raw as training data.

Each sequence folder in this release contains:

radar_*.h5                 raw radar samples and timestamps
camera_timestamps_*.h5     ZED timestamps
dji_timestamps_*.h5        DJI timestamps
max30105_*.h5              MAX30105 samples and timestamps (evaluation only)

Training sequences have no MAX30105 file. The processor handles that automatically.

Process a sequence

processing_code/processor.py matches the sensor timestamps and saves one output folder per sequence. processing_code/utils/ contains the sensor extractors, radar processing, synchronization, and point-cloud code. The processor also needs the ZED and DJI videos, so the public download alone cannot run it until the anonymized videos are released.

1. Install the ZED SDK on a Linux machine with a supported NVIDIA GPU.

2. From this repository's root folder, create a Python environment and install the dependencies:

   python3.12 -m venv .venv
   source .venv/bin/activate
   python -m pip install -r processing_code/requirements.txt
   python /usr/local/zed/get_python_api.py
   python -c 'import pyzed.sl as sl; print(sl.Camera.get_sdk_version())'
   

   Run the ZED Python API installer after activating the environment. See the Stereolabs instructions if your SDK uses another path. We tested the processor with Python 3.12 and ZED SDK 5.4.1; a different SDK version may produce slightly different ZED RGB/depth values.

3. Process one evaluation sequence:

   python processing_code/processor.py \
     --dataset GRADE_Eval_Raw \
     --sequences Smoke-keck-2-3 \
     --output-dir processed_eval
   

For a training sequence, use --dataset GRADE_Train_Raw --sequences AlicePratt-1. Leave out --sequences to process every folder in the chosen dataset.

The default output for one evaluation sequence looks like this:

processed_eval/Smoke-keck-2-3/
├── radar.npy            complex64 radar spectrum, (N, 64, 2, 8, 256)
├── dji_rgb.npy          uint8 RGB, (N, 504, 896, 3)
├── zed_rgb.npy          uint8 RGB, (N, 504, 896, 3)
├── zed_depth.npy        uint16 depth in millimeters, (N, 720, 1280)
├── max30105.npy         uint32, 3 channels; only when the sensor file exists
├── sync_triples.csv     original radar, ZED, and DJI frame indices
├── metadata.json        shapes, timestamps, and processing settings
└── pcd/
    ├── pcd_0.npy        float32 XYZ points, (number_of_points, 3)
    └── ...

Each point cloud is a separate .npy file because its number of points varies by frame. N is the number of synchronized frames kept for that sequence.

Useful options

Option	What it does
--dataset PATH	Choose GRADE_Train_Raw or GRADE_Eval_Raw (required).
--sequences NAME [NAME ...]	Process selected sequences; omit to process all.
--output-dir PATH	Choose where outputs go (default: processed).
--no-radar, --no-camera, --no-dji	Skip the named processed output. --no-radar skips the spectrum; use --no-pcd to skip point clouds. All three sensor timestamps are still used for synchronization.
--no-pcd	Skip point clouds (they are produced by default).
--no-doppler	Save radar_no_doppler.npy instead of the full radar.npy.
--list-unprocessed	List sequences without an output metadata.json.

--rgb-codec is a legacy argument and has no effect on the current .npy output. Run python processing_code/processor.py --help for the complete CLI help.

Please cite GRADE if you find this dataset useful in your research

@inproceedings{zhao2026grade,
  author    = {Bin Zhao and Patrick Chiou and Nakul Garg},
  title     = {{GRADE}: Single-Frame Generative Radar Depth Estimation Under Visual Degradation},
  booktitle = {Proceedings of the 32nd Annual International Conference on Mobile Computing and Networking},
  series    = {MobiCom '26},
  year      = {2026},
  publisher = {ACM},
  address   = {New York, NY, USA},
  numpages  = {15},
  doi       = {10.1145/3795866.3844478},
  url       = {https://doi.org/10.1145/3795866.3844478}
}