Deep-sea observatories images labeled by citizen for object detection algorithms
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Observatories provide continuous access to both coastal and deep-sea ecosystems, particularly from underwater imaging that is a non-destructive method for examining biodiversity on unprecedented time and space scales. The success of imagery data for scientific purposes leads to new challenges linked to the processing of the exponential amount of data collected, which can be time-consuming and tedious. Annotated images databases are generated by scientists, students, technical staff in laboratories, as well as by citizens through online platforms. They can be used to train machines -through AI models- for automatic processing of images collected by cameras at observatories underwater sites, identifying and analysing fauna and habitats for ecosystem monitoring purposes. In this case, we prepared the citizen science annotations from Deep Sea Spy as a training dataset for YoloV8. Indeed, Deep Sea Spy is a participative science platform launched in 2017, that provides access to images from EMSO-Azores and Ocean Networks Canada observatories for annotation purposes. We also used an expert annotated dataset for model validation. The archive includes: - An Images directory containing 3979 images from both observatories - The raw dataset containing 253323 annotations with 15 labeled classes from Deep Sea Spy : Alvinocaridid shrimp, Brittle star, Buccinoid snail, Bythograeid crab, Cataetyx fish, Chimera fish, Mussel bed, Polynoid worm, Polynoid worms, Pycnogonid (Sea spider), Spider crab, Tubicolous worm bed, Zoarcid fish, Microbial mat, Other fish - The cleaned dataset containing 14967 annotations with the Buccinidae and Bythograeidae classes - The expert dataset used for training validation of the Buccinid class - YoloV8 trained models on Buccinidae and Bythograeidae (.pt files) More information about data format, data cleaning and model training is available in the README file. The full pipeline is freely available on github.com/ai4os-hub/deep-species-detection
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it