Zasolar Unified Reviewall A
Mask R-CNN (ResNet-50 + FPN) checkpoint for rooftop solar installation footprint segmentation in South Africa. This is the unifiedreviewallA (UnifiedRA-A) detector from the ZAsolar project — the current production...
Source du modèle
- Type
- Other
- Taille du fichier
- 168 MB
- Version
- vBest model
- Mis à jour
- 1 nov. 2026
- Source
- Hugging Face
Description de la source
ZAsolar — unified_reviewall_A
Mask R-CNN (ResNet-50 + FPN) checkpoint for rooftop solar installation footprint
segmentation in South Africa. This is the unified_reviewall_A (UnifiedRA-A)
detector from the ZAsolar project —
the current production model for the Cape Town / Johannesburg rooftop-solar census.
Files
best_model.pth: trained PyTorch checkpoint (~169 MB)
Training Summary
- Task: installation-level rooftop solar footprint segmentation (reviewed-prediction footprints)
- Architecture: torchvision Mask R-CNN, ResNet-50-FPN backbone,
num_classes=2 - Warm-start: V3-C targeted hard-negative checkpoint (
exp003_C_targeted_hn) - Training data: Cape Town aerial_2025 + Johannesburg Vexcel 2024, per-instance mask-trusted halo gating
- Training set id:
unified_reviewall_20260511 - Trained: 2026-05-13
- Post-processing: per-detection finalize + SAM 2.1 mask+box refinement (
configs/postproc/v4_canonical.json)
Evaluation
Grid-aggregate installed-area inventory is the primary metric; per-polygon F1 is diagnostic.
Calibration — Cape Town, 16 held-out grids (in-domain, PV/non-PV classifier attached, 2026-06-08):
| Metric | Value |
|---|---|
| Grid-level R2 (detected vs. truth area) | 0.955 |
| Aggregate area-F1 | 0.742 |
| Area precision / recall | 0.64 / 0.88 |
| Per-grid dispersion (σ_Bw) | 0.260 |
| Coverage @ IoU 0.5 | 0.829 |
Validation — cross-domain, 6 unseen South African cities × 60 grids (Vexcel aerial, 2026-05-31):
| Metric | Value |
|---|---|
| Grid-level R2 | 0.971 |
| Aggregate area-F1 | 0.763 |
| Bulk ratio (detected / truth) | 1.04 |
| Area precision / recall | 0.75 / 0.78 |
Aggregate calibration holds on cities the model never trained on; per-grid dispersion widens out-of-domain (σ_Bw 0.26 → 1.23), which the downstream PV/non-PV classifier and per-city threshold calibration are designed to absorb.
Usage
from huggingface_hub import hf_hub_download
from core.models.maskrcnn import build_solar_maskrcnn # from the ZAsolar repo
ckpt = hf_hub_download("botao0818/zasolar-unified-reviewall-A", "best_model.pth")
model = build_solar_maskrcnn(pretrained_path=ckpt, num_classes=2).eval()
The checkpoint is a plain PyTorch state dict (raw, or nested under model/state_dict).
For the full inference + evaluation pipeline, see detect_and_evaluate.py in the repo.
Intended Use
Research and geospatial analysis around rooftop solar detection. Not a drop-in Transformers model; load it with the project's PyTorch code.
Limitations
- Trained for ZAsolar imagery and labeling conventions
- Evaluation conclusions depend on the repository's explicit
installationprofile and grid-aggregate framing - Out-of-domain generalization (new geographies / imagery sources) shows wider per-grid dispersion (see validation)
License
MIT (code/weights). Annotations governed separately — see repo data/annotations/ANNOTATION_SPEC.md.
Sources
1 source
Sources
1 sourceVérifié 1 nov.
Artefacts du modèle
1 artefact
Artefacts du modèle
1 artefactExtraits de sources
2 extraits
Extraits de sources
2 extraits--- license: mit library_name: pytorch pipeline_tag: image-segmentation tags: - solar - rooftop-solar - instance-segmentation - mask-rcnn - remote-sensing - geospatial --- # ZAsolar — unified_reviewall_A Mask R-CNN (ResNet-50 + FPN) checkpoint for rooftop solar installation footprint segmentation in South Africa. This is the `unified_reviewall_A` (UnifiedRA-A) detector from the [ZAsolar](https://github.com/Robertgao0818/ZAsolar) project — the current production model for the Cape Town / Johannesburg rooftop-solar census. ## Files - `best_model.pth`: trained PyTorch checkpoint (~169 MB) ## Training Summary - Task: installation-level rooftop solar footprint segmentation (reviewed-prediction footprints) - Architecture: torchvision Mask R-CNN, ResNet-50-FPN backbone, `num_classes=2` - Warm-start: V3-C targeted hard-negative checkpoint (`exp003_C_targeted_hn`) - Training data: Cape Town aerial_2025 + Johannesburg Vexcel 2024, per-instance mask-trusted halo gating - Training set id: `unified_reviewall_20260511` - Trained: 2026-05-13 - Post-processing: per-detection finalize + SAM 2.1 mask+box refinement (`configs/postproc/v4_canonical.json`) ## Evaluation Grid-aggregate installed-area inventory is the primary metric; per-polygon F1 is diagnostic. **Calibration — Cape Town, 16 held-out grids (in-domain, PV/non-PV classifier attached, 2026-06-08):** | Metric | Value | |---|---| | Grid-level R² (detected vs. truth area) | 0.955 | | Aggregate area-F1 | 0.742 | | Area precision / recall | 0.64 / 0.88 | | Per-grid dispersion (σ_Bw) | 0.260 | | Coverage @ IoU 0.5 | 0.829 | **Validation — cross-domain, 6 unseen South African cities × 60 grids (Vexcel aerial, 2026-05-31):** | Metric | Value | |---|---| | Grid-level R² | 0.971 | | Aggregate area-F1 | 0.763 | | Bulk ratio (detected / truth) | 1.04 | | Area precision / recall | 0.75 / 0.78 | Aggregate calibration holds on cities the model never trained on; per-grid dispersion widens out-of-domain (σ_Bw 0.26 → 1.23), which the downstream PV/non-PV classifier and per-city threshold calibration are designed to absorb. ## Usage ```python from huggingface_hub import hf_hub_download from core.models.maskrcnn import build_solar_maskrcnn # from the ZAsolar repo ckpt = hf_hub_download("botao0818/zasolar-unified-reviewall-A", "best_model.pth") model = build_solar_maskrcnn(pretrained_path=ckpt, num_classes=2).eval...
Source context: 0 downloads · 0 likes · Pipeline image-segmentation · Library pytorch · Repo botao0818/zasolar-unified-reviewall-A
