license: mit
task_categories:
- image-to-text
- image-to-3d
- image-feature-extraction
- text-to-3d
- feature-extraction
size_categories:
- 1K<n<10K
Dataset Card for HUVER
The dataset is comprised of a 6,051 unique UAV configurations, where each configuration is described by multiple data for- mats, including a grammar string, an RGB image, and an GLB file. Complementing these representation modalities, we also provide a configuration-based description, i.e., a text descriptor describing the features of each UAV using natural language
- Curated by: Abhiram Karri, Gary Stump, Christopher McComb, Binyang Song
- Language(s) (NLP): English, en
- License: MIT
Dataset Sources [optional]
- Repository: HUVER
- Paper [optional]: HUVER: The HyForm Uncrewed Vehicle Engineering Repository, ASME JMD (under review)
Uses
Direct Use
- The multimodal UAV dataset presented in this paper features a variety of UAV representations, including 3D models in GLB format, grammar representations, text descriptions, and parametric data. This diversity facilitates the development of surrogate models that utilize different UAV representations to predict performance more accurately.
- This multifaceted UAV dataset supports the creation of diverse UAV designs through various generative models. Models like GANs, LSTMs, transformers, and GNNs can generate new UAV images and designs in different formats, including GLB. The inclusion of negative examples in the dataset helps identify and correct potential design flaws early, enhancing model refinement and ensuring the feasibility and safety of the UAV designs.
Out-of-Scope Use
This dataset is not appropriate for detailed design of flight ready systems.
Dataset Structure
Data Instances
{
'Image': <0001.png>,
'glb_file': <https://huggingface.co/datasets/raiselab/HUVER/resolve/main/train/glb/0001.glb?download=true>,
'Grammar_string": <*aMM0-*bNM2++*cMN1++*dLM2*eML1^ab^ac^ad^ae>,
'Cost ($)': <1877.19>,
'Number of Batteries': <1>,
'Number of Motor-Rotor Pairs': <4>, 'Number of Airfoils': <0>,
'Number of Connectors': <4>,
'Weight of Batteries (lb)': <19.40347644>,
'Weight of Motor-Rotor Pair (lb)': <3.858051314>,
'Weight of Airfoils (lb)': <0.0>,
'Total Weight (lb)': <23.26152854>,
'Total Thrust (lb)': <82.50002518>,
'Normalized Average Structure Size': <0.324324324>,
'Normalized Average Motor Size': <0.259259259>,
'Normalized Average Foil Size': <0.0>,
'Design Descriptor': <This drone is made up of 1 part and has 4 engines that help it move. It also has 0 wings for better flying. It has 4 links that connect everything together securely. The drone weighs 23.2615285432816 pounds in total and can lift itself and more, thanks to its strong thrust of 82.500025177002 pounds.>,
'Operations Descriptor': <This drone configuration has a feasible flying range of 0.0-0.0 miles, evaluated over the payload range of 0-0 pounds. This configuration has a velocity range of 0.06352621-0.06352621 mph. It is observed when payload increases, the flying range and velocity decrease. The drones achieve highest values of velocity and range for the lowest payloads. It can be interpreted from the data that the drone can fly as far as 0.0 miles, and can reach maximum speeds up to 0.06352621 mph. This means that while the drone does well in many situations, how far and fast it can fly can vary with how much payload it carries. This drone costs around $1877.19, adding up costs of all the components used to achieve this configuration.>,
'Performance': <Feasibilty": "CouldNotStabilize", "Flying Range": 0.0, "Payload Capacity (lb)": 0, "Velocity (mph)": 0.06352621, "Performance Descriptor": "This drone could not hover. The drone for a payload of 0 pounds, could not accomplish a successful run, the reason being either the motors could not provide enough lift or the drone did not balance properly after flight".>
}
Data Fields
- Grammar String : Each UAV configurationn in this dataset can be fully described by a grammar string, which is structured according to specific pre-defined grammar rules.
- Image : Top-down view of RGB Image of UAV corresponding to the UAV configuration (grammar string).
- glb : 3D mesh representation of the detailed spatial structure of the corresponding UAV configuration.
- Configuration parameter fields : Number of Batteries, Number of Motor-Rotor Pairs, Number of Airfoils, Number of Connectors, Weight of Batteries (lb), Weight of Motor-Rotor Pair (lb), Weight of Airfoils (lb), Total Weight (lb), Total Thrust (lb), Normalized Average Structure Size, Normalized Average Motor Size, Normalized Average Foil Size.
- Text Descriptions: Design Descriptor, provides a design description based on a UAV configuration. Performance Descriptor, introduces a description of the performance of a UAV based on its simulation results. Operational Descriptor, offers a description of the performance curve of a UAV based on its operation range
Visualization
Once you've downloaded the dataset you can visualize it using this script.
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data
Data Collection and Processing
[More Information Needed]
Who are the source data producers?
[More Information Needed]
Annotation process
[More Information Needed]
Who are the annotators?
[More Information Needed]
Personal and Sensitive Information
[More Information Needed]
Bias, Risks, and Limitations
[More Information Needed]
Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.