DOI: 10.17596/0004035
How to Cite
The following is an example of data citation including creator(s), publication year, title, publisher, DOI and access date.
Nishio, T., & Kawae, Y. (2025). An Image AI Dataset for Deep-sea Organism Detection. JAMSTEC. https://doi.org/10.17596/0004035
Abstract
This is an AI training dataset for detecting organisms from images taken during deep-sea survey. The detection targets are labeled into 19 rough categories such as "shrimp" and "fish". The dataset consists of over 8,000 images sliced from JAMSTEC's deep-sea survey videos on "JAMSTEC E-library of Deep-sea Images (J-EDI)" and COCO (Common Objects in Context) format JSON including bounding boxes and labels. The dataset can be downloaded as a zip file.
How to Access
| English | https://www.godac.jamstec.go.jp/jedi/e/dataset/jedi_organism_detection_dataset.html | |
| Japanese | https://www.godac.jamstec.go.jp/jedi/j/dataset/jedi_organism_detection_dataset.html |
Related Information
Link to CiNii Research
Full Metadata
| Title | An Image AI Dataset for Deep-sea Organism Detection |
| Creator(s) | Takaki Nishio [https://orcid.org/0009-0006-1452-9687] Yuki Kawae |
| Publication year | 2025 |
| Publisher | JAMSTEC |
| Resource type | Dataset/Scientific data |
| Right(s) | The copyright of the dataset belongs to JAMSTEC.Please refer to JAMSTEC's web page (https://www.jamstec.go.jp/e/database/data_policy.html) for details of the data policy. The dataset is licensed under CC BY-NC-SA 4.0.Please refer to the dataset download page (https://www.godac.jamstec.go.jp/jedi/e/dataset/jedi_organism_detection_dataset.html) for details of the conditions of use. JAMSTEC Data Policy [https://www.jamstec.go.jp/e/database/data_policy.html] CC BY-NC-SA 4.0 [https://creativecommons.org/licenses/by-nc-sa/4.0/] |
| Description | [Abstract] This is an AI training dataset for detecting organisms from images taken during deep-sea survey. The detection targets are labeled into 19 rough categories such as "shrimp" and "fish". The dataset consists of over 8,000 images sliced from JAMSTEC's deep-sea survey videos on "JAMSTEC E-library of Deep-sea Images (J-EDI)" and COCO (Common Objects in Context) format JSON including bounding boxes and labels. The dataset can be downloaded as a zip file. |
DOI Status
| Last Update | 2026-03-27 |
| Registered | 2025-03-05 |