Highly precise community science annotations of video camera‐trapped fauna in challenging environments
Bibliographic record
Abstract
Abstract As camera trapping grows in popularity and application, some analytical limitations persist including processing time and accuracy of data annotation. Typically images are recorded by camera traps although videos are becoming increasingly collected even though they require much more time for annotation. To overcome limitations with image annotation, camera trap studies are increasingly linked to community science (CS) platforms. Here, we extend previous work on CS image annotations to camera trap videos from a challenging environment; a dense tropical forest with low visibility and high occlusion due to thick canopy cover and bushy undergrowth at the camera level. Using the CS platform Chimp&See, established for classification of 599 956 video clips from tropical Africa, we assess annotation precision and accuracy by comparing classification of 13 531 1‐min video clips by a professional ecologist (PE) with output from 1744 registered, as well as unregistered, Chimp&See community scientists. We considered 29 classification categories, including 17 species and 12 higher‐level categories, in which phenotypically similar species were grouped. Overall, annotation precision was 95.4%, which increased to 98.2% when aggregating similar species groups together. Our findings demonstrate the competence of community scientists working with camera trap videos from even challenging environments and hold great promise for future studies on animal behaviour, species interaction dynamics and population monitoring.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".