Classifying Tracked Objects and their Interactions from Infrared Imagery
Bibliographic record
Abstract
For many intelligent security systems the use of infrared technology is becoming essential and is a challenging issue. This paper outlines a framework for exploiting spatio-temporal tracking parameters to classify multiple moving objects and recognize their interactions using low quality thermal imagery. For outdoor scenes, motion segmentation is automatically performed using a novel dynamic background-subtraction technique which robustly adapts detection to illumination changes. During the tracking process, the algorithm uses thermal data and motion parameters to label moving objects into two main classes: persons and vehicles. However, the calibrated temperature informations are employed to locate the "part of interest" of each classified object: head's parts for persons and engine's part for vehicles. Once these tasks are correctly performed, the algorithm applies predefined rules to recognize some interactions between classified objects. The reasoning rules are based on the class's labels and motion parameters to partition interactions according to their authors: events with a single author and events with multiple authors. For each of them, several actions of interest can be identified by testing specific criterion's combinations. This partition is the base of our algorithmic reasoning and increases considerably the process's speed. Thus, with this complete architecture, the system is able to automatically obtain relevant video surveillance scenarios useful as powerful tools for the decision-making aid and also for archiving. Finally, experiments proved that the algorithm is efficient and fast enough to operate in real-time implementation
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 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) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".