14,492 annotated underwater images form a large-scale dataset for training AI models in marine biodiversity research. The dataset is designed for multi-label image classification tasks and was created by lombardata. It was last updated on the platform in April 2025.
Use Cases
- Multi-label classification of marine species based on annotated underwater images
- Training object detection models for coral reef conservation research
- Benchmarking AI model performance on citizen-sourced ecological imagery
Strengths
- 14,492 annotated images provide a substantial base for model training
- Explicitly designed for multi-label classification tasks in a defined domain
Limitations
- Column-level documentation is absent; field semantics must be inferred after download
- Row count is unknown, which may limit suitability assessment
Provenance
- Source
- lombardata
- Collection Method
- Likely contains images contributed through citizen science initiatives.
- Time Range
- null
- Freshness
- Last updated 2025-04-29 05:41:50
- Geography
- null