license: mit
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: hsi
dtype: image
- name: rgb
dtype: image
- name: segmentation
dtype: image
- name: spectra
dtype:
array2_d:
shape:
- 1
- 305
dtype: float32
splits:
- name: train
num_bytes: 58018135163
num_examples: 504
download_size: 56572584220
dataset_size: 58018135163
task_categories:
- image-classification
- image-segmentation
language:
- en
tags:
- Hyperspectral
- Robotics
- Robot
- Autonomous
- Multispectral
pretty_name: 'Hyper-Drive: Hyperspectral Driving Dataset'
size_categories:
- n<1K
Towards automated analysis of large environments, hyperspectral sensors must be adapted into a format where they can be operated from mobile robots. In this dataset, we highlight hyperspectral datacubes collected from the Hyper-Drive imaging system. Our system collects and registers datacubes spanning the visible to shortwave infrared (660-1700 nm) in 33 wavelength channels. The system also simultaneously captures the ambient solar spectrum reflected off a white reference tile. The dataset consists of 500 labeled datacubes from on-road and off-road terrain compliant with the ATLAS.
This work first appeared at WHISPERS 2023 in Athens, Greece and is a product of the Robotics and Intelligent Vehicles Research Laboratory (RIVeR) at Northeastern University.
The paper is is available on Arxiv. If you use our data in your research, please make sure use the following citation:
@inproceedings{hanson2023hyper,
title={Hyper-Drive: Visible-Short Wave Infrared Hyperspectral Imaging Datasets for Robots in Unstructured Environments},
author={Hanson, Nathaniel and Pyatski, Benjamin and Hibbard, Samuel and DiMarzio, Charles and Pad{\i}r, Ta{\c{s}}k{\i}n},
booktitle={2023 13th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS)},
pages={1--5},
year={2023},
organization={IEEE}
}
Sample Dataloading Code
Hugging Face by default has difficulty loading multi-channel TIFFs. This snippet shows an example of how to load the dataset. It is highly recommended to enable streaming through the Hugging Face API.
from datasets import load_dataset, Image
from io import BytesIO
import tifffile
def decode_hyperspectral_tiff(batch):
batch["hsi"] = tifffile.imread(BytesIO(batch['hsi']["bytes"]))
return batch
dataset = load_dataset("nhanson2/hyper_drive", split='train', streaming=True)
dataset = dataset.cast_column("rgb", Image(mode='RGB', decode=True))
dataset = dataset.cast_column("hsi", Image(decode=False))
dataset = dataset.cast_column("segmentation", Image(mode='L', decode=True))
dataset = dataset.map(decode_hyperspectral_tiff)
In addition to the 500 labeled hyperspectral datacubes, raw ROS bagfiles generated of each of the sensor feeds at a higher frame rate are available here. These files are provided as an additional resource and do not contain semantic labels, but contain ~10,000 additional hyperspectral datacubes of in-between frames from the labeled dataset. It also contains additional datatypes for terrain analysis such as inertial measurement unit (IMU) data. To the best of the authors knowledge, it is the largest vehicle-centric hyperspectral dataset currently available!