Advanced Underwater Object Detection Images for Model Training

M. Aswani Kumar introduced a lightweight underwater object detector on figshare. The 2.7 GB dataset is licensed CC-BY-4.0 and was last updated on 2026-05-28. It is aimed at solving inherent problems in underwater images.

Use Cases

• Training object detection models based on images with spectral distortion.
• Benchmarking model performance on low-visibility underwater scenes.
• Developing lightweight neural networks for real-time underwater detection tasks.
• Studying model robustness against background clutter and object ambiguity.

Strengths

• The dataset is 2.7 GB, indicating a medium-scale collection of image data.
• It is specifically designed to address known challenges in underwater computer vision.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Freshness should be verified as the last update date is in the future (2026-05-28).

Provenance

Source

M. Aswani Kumar via figshare

Freshness

Last updated 2026-05-28 07:29:36