CoCoaSpec: Hyperspectral Images of Colombian Cocoa Beans with Physicochemical Labels

CoCoaSpec is a multimodal hyperspectral imaging dataset of Colombian cocoa beans with detailed physicochemical annotations. It was created by ecos-nord-ginp-uis to support research on non-destructive quality assessment, spectral data analysis, and multimodal data fusion. The dataset includes hyperspectral images acquired with four different devices and was last updated on November 10, 2025.

Use Cases

• Train models for non-destructive cocoa quality assessment based on hyperspectral images and physicochemical annotations.
• Develop spectral data analysis methods using images from four different acquisition devices.
• Research multimodal data fusion techniques combining hyperspectral imagery with lab-measured physical and chemical properties.

Strengths

• Dataset includes hyperspectral images from four different acquisition devices.
• Contains detailed physicochemical annotations for cocoa beans.
• Explicitly designed for research on non-destructive quality assessment and multimodal fusion.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Data may reflect geographic bias inherent to its collection from Colombian cocoa beans.

Provenance

Source

huggingface

Collection Method

Hyperspectral imaging and physicochemical lab analysis.

Freshness

Last updated 2025-11-10 16:16:30.

Geography

Colombia