|
Download docs/experiments.md from fawad-archAI/building-defect-ai: direct link, hf CLI and curl.
- Browser
- Download file 1.43 kB
-
https://huggingface.co/fawad-archAI/building-defect-ai/resolve/main/docs/experiments.md
- Command line
-
hf download hf://fawad-archAI/building-defect-ai/docs/experiments.md
-
curl -L -o experiments.md https://huggingface.co/fawad-archAI/building-defect-ai/resolve/main/docs/experiments.md
1.43 kB
| # Experiment Results | |
| ## Model comparison | |
| | Experiment | Training strategy | Test accuracy | | |
| |---|---|---:| | |
| | Experiment 1 | Frozen ResNet18 feature extractor; train classifier | 83.14% | | |
| | Experiment 2 | Fine-tune Layer3, Layer4 and classifier | 79.75% | | |
| | Experiment 3 | Fine-tune Layer4 and classifier with early stopping | **87.92%** | | |
| ## Experiment 3 — Selected model | |
| Experiment 3 achieved a test accuracy of **87.92% across 795 test images**. | |
| The model used selective fine-tuning of ResNet18's Layer4 and final classification layer. Early stopping was applied using validation loss to limit unnecessary training. | |
| The saved checkpoint was: | |
| `models/building_defect_resnet18_experiment3.pth` | |
| ## Comparison and interpretation | |
| Experiment 3 outperformed the first two experiments on the recorded test evaluation. | |
| Experiment 2 showed a larger training-validation performance gap, suggesting overfitting. Experiment 3 used a more constrained fine-tuning strategy and achieved the strongest recorded test accuracy. | |
| The confusion matrix also indicates that some defect categories remain difficult to distinguish, particularly the crack categories. | |
| ## Conclusion | |
| Experiment 3 is selected as the project's final candidate model based on its measured test accuracy and training strategy. | |
| Further work should investigate class-specific errors, dataset diversity and possible near-duplicate images across dataset splits. |