YOLOv7 for Real-Time Monitoring of Heat Loss Regions in Building Envelopes Using UAV-Infrared Thermography
Bibliographic record
Abstract
Drone technology has enabled efficient city-scale data collection for detecting thermal bridges in building envelopes. These thermal bridges can cause up to 30% of energy loss and can make up to 50% of the building's envelope area leading to severe heat loss, moisture infiltration, and mold growth. Addressing thermal bridges through timely monitoring can enhance building energy performance and offer comfortable occupied conditions. This paper offers a rapid, non-invasive solution for identifying areas of heat loss to improve building energy efficiency. The study utilized real world thermal bridging data collected from six city building blocks, each block containing approximately 20 buildings. The data was collected using a DJI M600 drone equipped with a FLIR-XT2 thermal camera, capturing thermal images of building envelopes. The YOLOv7 deep learning model was used to process the collected data for automated thermal anomaly detection. The model was trained over 250 epochs using standard gradient descent algorithm. During the training process, the loss function decreased gradually. The model achieved a precision of 0.58 in detecting thermal bridges. The Drone-based approach allowed capturing images from about 20 different angles per building, enhancing detection reliability. Moreover, the model achieved a high inference speed of 41 frames per second at an image size of${640}\times {640}\ \text{pixels}$, providing a method for real-time data analysis of building façades, The results demonstrate that integrating thermal imaging with advanced deep learning techniques effectively detects thermal anomalies in buildings envelopes.
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 | 0.000 |
| 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 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 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".