Global DEMs to tackle RPC biases and the overfitting phenomenon in high-resolution satellite imagery
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The overfitting phenomenon and rational polynomial coefficients (RPCs) biases are two crucial issues that degrade the accuracy of geospatial products derived from high-resolution satellite images. The overfitting phenomenon is caused by both a large number of RPCs and strong correlations among them. The RPC biases arise from uncertainties in the global positioning system receivers and inertial measurement units. In this article, an innovative framework based on the genetic algorithm (GA) and the least squares (LS) algorithm, called GALS, is proposed to overcome these problems simultaneously. In this method, the GA is applied to select the optimum RPCs, while the LS algorithm is used to estimate the values of the optimally selected RPCs. The GALS method requires various sets of well-distributed ground control points (GCPs). To tackle the problem of GCP collection, we generated a large number of digital elevation model (DEM)-derived GCPs (DEMGCPs), using a global DEM (GDEM) and vendor-provided RPCs, refined by only one GCP. To evaluate the performance of this framework, four IRS-P5 data sets were used. The GALS is compared to two competing methods, L1-norm-regularized LS and ridge estimation by considering two scenarios using 50 GCPs and the DEMGCPs. The results demonstrate the superiority of GALS in both scenarios. Furthermore, GALS using DEMGCPs led to far more accurate and stable results when compared to GALS using GCPs. Compared to the vendor-provided RPCs, the results of the GALS using DEMGCPs also indicate a major improvement, single-pixel or subpixel accuracy with around 15 RPCs, and only 1 GCP, in both accuracy and reliability of georeferencing for all IRS-P5 data sets.
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.
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 | 0.001 | 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) | 0.000 | 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 it