A Novel Approach for Fall Detection Using Thermal Imaging and a Stacking Ensemble of Autoencoder and 3D-CNN Models
Bibliographic record
Abstract
Falls are a significant cause of injury and mortality, particularly among the elderly population. Early detection of falls is crucial to mitigate their impact. Thermal imaging is a promising technology for detecting falls as it is non-invasive and operates in low-light conditions. However, accurately detecting falls in thermal images remains challenging due to the low resolution and lack of color information in these images. This paper proposes a novel approach for improving fall detection in thermal image data using a stacking ensemble of Autoencoder (AE) and 3-D Convolutional Neural Network (3D-CNN) models fed into a meta-neural network which is trained to detect falls and non-falls. The effectiveness of the proposed system is demonstrated through ablation studies on the publicly available benchmark dataset—"Thermal Simulated Fall", in which the model achieves an accuracy of 83%. Comparative analysis shows that the proposed solution outperforms an AE-based baseline by 9.2%. The combination of AEs and 3D-CNNs allows us to harness the power of both supervised and unsupervised learning methodologies, whilst mitigating the limitations and biases of each individual model, offering a promising solution for accurate and efficient fall detection in thermal image input streams.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".