---
pretty_name: >-
  CoCoaSpec: A multimodal hyperspectral dataset of cocoa beans with physicochemical annotation
license: cc-by-4.0
authors:
  - Kebin Contreras
  - Mohamad Jouni
  - Mauro Dalla Mura
  - Jorge Bacca
---

# CoCoaSpec: A Multimodal hyperspectral dataset of cocoa beans with physicochemical annotation

## Overview
The **CoCoaSpec dataset** is a multimodal hyperspectral imaging dataset of Colombian cocoa beans with detailed physicochemical annotations.  
It was created to support research on **non-destructive cocoa quality assessment**, **spectral data analysis**, and **multimodal data fusion**.  

The dataset includes hyperspectral images acquired with four different devices, along with reference physicochemical measurements and metadata.

## Contents
- **Hyperspectral cubes** (raw and preprocessed)  
- **RGB images** (EOS M50 camera)  
- **Physicochemical annotations** (fermentation degree, moisture content, etc.)  
- **Calibration & metadata** (dark/flat fields, wavelength centers, camera metadata, acquisition conditions, calibration details, sample identifiers)

## Data Structure
The dataset is organized as follows:
```
data/
├── scenes/                 # Scene-level acquisitions across devices
├── resources/              # Calibration and metadata resources
│ ├── dark_fields/
│ ├── flat_fields/
│ ├── metadata/             # cameras.json, campaign_metadata.json
│ ├── wavelengths/          # per-device band centers
│ └── physicochemical.csv   # physicochemical information
├── README.md
└── dataset.zip                # Full dataset as a single archive
```

## How to Use
You can load the dataset with the Hugging Face `datasets` library:

```python
from datasets import load_dataset

# Login using e.g. `huggingface-cli login` to access this dataset
ds = load_dataset("ecos-nord-ginp-uis/CoCoaSpec")
```

## Code (Loading, Preprocessing, Visualization)
Example Python scripts for loading, visualization, and preprocessing are available in the public GitHub repository:  
https://github.com/kebincontreras/CoCoaSpec

## License
This dataset is released under the [Creative Commons Attribution 4.0 International (CC BY 4.0)](https://creativecommons.org/licenses/by/4.0/) license.

## Citation

If you use this dataset, please cite the **DOI** below.

### Nature Scientific Data style reference

> Contreras, K., Jouni, M., Dalla Mura, M. & Bacca, J.
> *CoCoaSpec: A multimodal hyperspectral dataset of cocoa beans with physicochemical annotation.*
> Hugging Face Datasets [https://doi.org/10.57967/hf/6961](https://doi.org/10.57967/hf/6961) (2025). *(Revision a6bf0d7)*

### BibTeX

```bibtex
@dataset{contreras2025cocoaspec,
  author    = {Contreras, Kebin and Jouni, Mohamad and Dalla Mura, Mauro and Bacca, Jorge},
  title     = {CoCoaSpec: A Multimodal hyperspectral dataset of cocoa beans with physicochemical annotation},
  year      = {2025},
  publisher = {Hugging Face Datasets},
  doi       = {10.57967/hf/6961},
  url       = {https://doi.org/10.57967/hf/6961},
  note      = {Revision a6bf0d7}
}
```

## Acknowledgements
This dataset was developed at Universidad Industrial de Santander (Colombia) in collaboration with Université Grenoble Alpes – GIPSA-Lab (France).  
We thank all contributors for their efforts in acquisition, annotation, and validation.

## Contact

For questions, suggestions, or issues regarding this dataset, please contact (primary first):

- **Kebin Contreras** — Universidad Industrial de Santander (UIS)  
  Email: [kebinandrescontreras@gmail.com](mailto:kebinandrescontreras@gmail.com?subject=CoCoaSpec%20dataset)

- **Mohamad Jouni** — Université Grenoble Alpes (UGA), GIPSA-Lab  
  Email: [mohamad.jouni@grenoble-inp.fr](mailto:mohamad.jouni@grenoble-inp.fr?subject=CoCoaSpec%20dataset)

- **Mauro Dalla Mura** — Grenoble INP–UGA, GIPSA-Lab  
  Email: [mauro.dalla-mura@gipsa-lab.grenoble-inp.fr](mailto:mauro.dalla-mura@gipsa-lab.grenoble-inp.fr?subject=CoCoaSpec%20dataset)

- **Jorge Bacca** — Universidad Industrial de Santander (UIS)  
  Email: [Jbacquin@uis.edu.co](mailto:Jbacquin@uis.edu.co?subject=CoCoaSpec%20dataset)

Please mention **“CoCoaSpec dataset”** in the subject line when reaching out.