Thermal camera data annotation for the Seeing Through the Fog dataset
Bibliographic record
Abstract
This dataset contains manually annotated thermal frames from the Seeing Through the Fog dataset (https://light.princeton.edu/datasets/automated_driving_dataset/), with labeling performed only on the test set based on the sequence_based_split.json file, which contains the weather conditions and file names associated to each weather. The annotations follow the YOLO format, where each image has a corresponding .txt file containing the class ID and normalized bounding-box coordinates. The labeled classes include Car, Pedestrian, Large Vehicle, and Ridable Vehicle. Some images have very poor visibility or no detectable objects at all. As a result, not all frames contain visible objects. For those cases, the label file is intentionally left empty or is not created. YOLO-based loaders naturally interpret missing or empty label files as images without annotations. The thermal test set consists of 1,142 frames distributed across different weather conditions. Specifically, there are 128 frames each for Clear Day, Clear Night, Snow Day, Snow Night, Light Fog Day, Light Fog Night, Dense Fog Day, and Dense Fog Night. Rain conditions include 54 frames for Rain (Day) and 64 frames for Rain (Night). The variation in the number of frames across weather conditions reflects the original dataset’s availability and the differences in recorded sequences. This labeled subset is designed for evaluating object detection performance on thermal imagery under diverse and challenging weather conditions, and for analyzing how visibility variations affect detection robustness.
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 | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.004 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.015 | 0.030 |
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".