Enhancing Aerial Data Semantic Segmentation with a Colour Range Mask Layer: A Deep Learning Approach
Bibliographic record
Abstract
Recent advancements in airborne platforms equipped with ultra-high-resolution imaging sensors have significantly improved our capability to acquire detailed urban optical imagery.These systems offer exceptional capabilities for capturing highly precise and detailed urban data, paving the way for the generation of high-definition maps (HD maps) for innovative urban applications.However, manually extracting information from this data is a generally slow and labour-intensive process.Thus, employing deep learning algorithms for data extraction in such a context might be an alternative solution.Deep learning has revolutionised and transformed remote sensing and image analysis, especially in semantic segmentation, which divides images into meaningful regions.This transformative power of deep learning is particularly significant in urban analysis (e.g., urban planning, navigation, disaster management, and monitoring infrastructure), where detailed spatial information is crucial.Even though deep learning offers excellent potential, applying deep learning for semantic segmentation of images from urban environments presents several challenges.First, supervised deep learning algorithms require many training data to work effectively.Second, training and analysing ultrahigh-resolution (less than 5 cm) images with deep learning algorithms need large storage capacity, are computationally intensive and often require advanced data augmentation, pre-processing, and model optimisation techniques to achieve optimal results Zhu et al., (2017).
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 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".