Leveraging Thermal Infrared Imaging for Pig Ear Detection Research: The TIRPigEar Dataset and Performances of Deep Learning Models
Bibliographic record
Abstract
The stable physiological structure and rich vascular network of pig ears contribute to distinct thermal characteristics, which can reflect temperature variations. While the temperature of the pig ear does not directly represent core body temperature due to the ear's role in thermoregulation, thermal infrared imaging offers a feasible approach to analyzing individual pig status. Based on this background, a dataset comprising 23,189 thermal infrared images of pig ears (TIRPigEar) was established. The TIRPigEar dataset was obtained through a pig house inspection robot equipped with an infrared thermal imaging device, with post-processing conducted via manual annotation. By labeling pig ears within these images, a total of 69,567 labeled files were generated, which can be directly used for training pig ear detection models and enabling the analysis of pig temperature information by integrating the corresponding thermal imaging data. To validate the dataset's utility, it was evaluated across various object detection algorithms. Experimental results show that the dataset achieves the highest precision, recall, and mAP50 on the YOLOv9m model, reaching 97.35%, 98.1%, and 98.6%, respectively. Overall, the TIRPigEar dataset demonstrates optimal performance when applied to the YOLOv9m algorithm. Utilizing thermal infrared imaging technology to detect pig ear information provides a non-contact, rapid, and effective method. Establishing the TIRPigEar dataset is highly significant, as it allows for a valuable resource for AI and precision livestock farming researchers to validate and improve their algorithms. This dataset will support many researchers in advancing precision livestock farming by enabling an efficient way for pig ear temperature analysis.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.003 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.003 |
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".