Initial upload: 5-fold CV ensemble checkpoints (Kaggle test 0.84690)

README.md

@@ -1,3 +1,92 @@
 ---
 license: mit
+tags:
+  - image-segmentation
+  - historical-maps
+  - cadastral-maps
+  - block-vectorization
+  - cross-validation
+  - ensemble
+library_name: pytorch
+pipeline_tag: image-segmentation
 ---
+
+# Historical Map City-Block Vectorisation — 5-fold CV ensemble
+
+EfficientNet-B4 UNet + SCSE attention checkpoints from the 5-fold cross-validation ensemble used for the Research Topics in Cartography (RTCart) 2026 Task 2 competition. Together they constitute the model that scored **0.84690** on the Kaggle leaderboard (`score = 0.4 × c-IoU + 0.6 × c-PoLiS`).
+
+The corresponding code lives in the **[NB11/block_vectorization](https://github.com/NB11/block_vectorization)** repository — clone that to actually run inference.
+
+## What's here
+
+| Folder | Fold | Val IoU (held-out tiles) |
+| --- | --- | --- |
+| `breezy-yogurt-59/`     | 0 | 0.9845 |
+| `autumn-lake-60/`       | 1 | 0.9852 |
+| `legendary-deluge-61/`  | 2 | 0.9870 |
+| `serene-durian-62/`     | 3 | 0.9824 |
+| `worthy-blaze-63/`      | 4 | 0.9826 |
+
+Each folder contains:
+- `checkpoint.pth` — model state dict (~80 MB)
+- `config.yaml` — exact training config
+
+Plus `cv5_manifest.json` — fold index → folder mapping for the inference orchestrator.
+
+## How to use
+
+```bash
+# 1. Clone the code repo
+git clone https://github.com/NB11/block_vectorization.git
+cd block_vectorization
+python -m venv .venv && source .venv/bin/activate
+pip install -r requirements.txt
+
+# 2. Download these weights into the runs/ tree
+mkdir -p runs
+for fold in breezy-yogurt-59 autumn-lake-60 legendary-deluge-61 serene-durian-62 worthy-blaze-63; do
+  mkdir -p "runs/$fold"
+  curl -L "https://huggingface.co/Noe-B/historical-map-city-block-vectorization/resolve/main/$fold/checkpoint.pth" \
+       -o "runs/$fold/checkpoint.pth"
+  curl -L "https://huggingface.co/Noe-B/historical-map-city-block-vectorization/resolve/main/$fold/config.yaml" \
+       -o "runs/$fold/config.yaml"
+done
+curl -L "https://huggingface.co/Noe-B/historical-map-city-block-vectorization/resolve/main/cv5_manifest.json" \
+     -o "runs/cv5_manifest.json"
+
+# 3. Generate the data manifest + run the ensemble (requires the raw maps in data/raw/)
+python scripts/pipeline/1_preprocess.py
+python scripts/pipeline/1b_make_cv_manifest.py --block-size 2 --n-folds 5
+python scripts/pipeline/3b_infer_cv.py $(jq -r '.folds[]' runs/cv5_manifest.json)
+python scripts/pipeline/5_postprocess.py runs/cv5-breezy-yogurt-59 config/postprocess_cv_optimal.yaml
+python scripts/pipeline/6_submit.py      runs/cv5-breezy-yogurt-59 config/postprocess_cv_optimal.yaml
+```
+
+## Model
+
+- **Architecture:** `segmentation_models_pytorch` UNet decoder with SCSE attention, EfficientNet-B4 encoder (ImageNet-pretrained), 768 × 768 input, 2-channel output (interior + boundary ring).
+- **Loss:** `BceLovász` on interior + `BCEDice` on boundary ring.
+- **Training:** AdamW + cosine LR schedule, batch_size 2, early-stopping patience 30, max 120 epochs. All 5 folds warm-started from the Stage 2 anchor; each was fine-tuned at LR 4e-5 on its held-out fold split.
+- **Data:** ~190 tiles per fold (the other ~48 held out for validation), spanning both labelled maps. 5-fold spatial-block stratified CV (block_size_tiles = 2) prevents train/val leakage.
+
+## Results
+
+| Stage | Approach | Kaggle test |
+| --- | --- | --- |
+| Single model (Stage 3) | `fresh-sunset-57` (boundary head, single-map PP sweep) | 0.840 |
+| 5-fold CV ensemble | Same PP as Stage 3 | 0.83686 |
+| **5-fold CV + CV-aware PP** | this release | **0.84690** |
+
+The CV-aware postprocessing config that produced the final score is committed in the code repo at `config/postprocess_cv_optimal.yaml`.
+
+## Citation
+
+```bibtex
+@misc{brugger2026blockvec,
+  title  = {Block-Polygon Extraction from Historical Cadastral Maps},
+  author = {Brugger, Noé},
+  year   = {2026},
+  note   = {RTCart 2026, Task 2},
+  url    = {https://github.com/NB11/block_vectorization}
+}
+```

autumn-lake-60/checkpoint.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f737a32b0d764e708cdab18ef5f03515ca089fbe9eb20256aba876f6036a0452
+size 82028033

autumn-lake-60/config.yaml

@@ -0,0 +1,33 @@
+experiment_name: cv_fold1
+preprocessing:
+  approach: v2
+  manifest_path: data/preprocessed/v2_manifest_cv.json
+model:
+  name: efficientnet_unet
+  encoder: efficientnet-b4
+  encoder_weights: imagenet
+  attention: scse
+  in_channels: 3
+  classes: 2
+  tile_size: 768
+training:
+  epochs: 120
+  batch_size: 2
+  learning_rate: 4.0e-05
+  optimizer: adamw
+  scheduler: cosine
+  loss: boundary_head
+  augmentation: true
+  num_workers: 2
+  n_folds: 5
+  fold_index: 1
+  early_stopping_patience: 30
+  early_stopping_min_delta: 0.0001
+  init_from: runs/civilized-wars-52/checkpoint.pth
+loss:
+  boundary_weight: 0.5
+  bce_weight: 0.5
+  lovasz_weight: 0.5
+wandb:
+  project: block-vectorization
+  entity: research_carto

breezy-yogurt-59/checkpoint.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3afa3aae695f26aabaee6165695cbc132fb240822eb581cf35cbed3f28afa393
+size 82028033

breezy-yogurt-59/config.yaml

@@ -0,0 +1,33 @@
+experiment_name: cv_fold0
+preprocessing:
+  approach: v2
+  manifest_path: data/preprocessed/v2_manifest_cv.json
+model:
+  name: efficientnet_unet
+  encoder: efficientnet-b4
+  encoder_weights: imagenet
+  attention: scse
+  in_channels: 3
+  classes: 2
+  tile_size: 768
+training:
+  epochs: 120
+  batch_size: 2
+  learning_rate: 4.0e-05
+  optimizer: adamw
+  scheduler: cosine
+  loss: boundary_head
+  augmentation: true
+  num_workers: 2
+  n_folds: 5
+  fold_index: 0
+  early_stopping_patience: 30
+  early_stopping_min_delta: 0.0001
+  init_from: runs/civilized-wars-52/checkpoint.pth
+loss:
+  boundary_weight: 0.5
+  bce_weight: 0.5
+  lovasz_weight: 0.5
+wandb:
+  project: block-vectorization
+  entity: research_carto

cv5_manifest.json

@@ -0,0 +1,12 @@
+{
+  "started_at": "2026-05-19T23:21:51.460456",
+  "n_folds": 5,
+  "folds": {
+    "0": "runs/breezy-yogurt-59",
+    "1": "runs/autumn-lake-60",
+    "2": "runs/legendary-deluge-61",
+    "3": "runs/serene-durian-62",
+    "4": "runs/worthy-blaze-63"
+  },
+  "last_updated": "2026-05-20T05:06:55.390755"
+}

legendary-deluge-61/checkpoint.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b591d597e453d2ed3b05dc09c47bee464fab81905cbafcdaa9ecf3a7627a3ab9
+size 82028033

legendary-deluge-61/config.yaml

@@ -0,0 +1,33 @@
+experiment_name: cv_fold2
+preprocessing:
+  approach: v2
+  manifest_path: data/preprocessed/v2_manifest_cv.json
+model:
+  name: efficientnet_unet
+  encoder: efficientnet-b4
+  encoder_weights: imagenet
+  attention: scse
+  in_channels: 3
+  classes: 2
+  tile_size: 768
+training:
+  epochs: 120
+  batch_size: 2
+  learning_rate: 4.0e-05
+  optimizer: adamw
+  scheduler: cosine
+  loss: boundary_head
+  augmentation: true
+  num_workers: 2
+  n_folds: 5
+  fold_index: 2
+  early_stopping_patience: 30
+  early_stopping_min_delta: 0.0001
+  init_from: runs/civilized-wars-52/checkpoint.pth
+loss:
+  boundary_weight: 0.5
+  bce_weight: 0.5
+  lovasz_weight: 0.5
+wandb:
+  project: block-vectorization
+  entity: research_carto

serene-durian-62/checkpoint.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1976d81524694b93724a298d80630e2adb027662b1a4ab68b4e19ee8b907334
+size 82028033

serene-durian-62/config.yaml

@@ -0,0 +1,33 @@
+experiment_name: cv_fold3
+preprocessing:
+  approach: v2
+  manifest_path: data/preprocessed/v2_manifest_cv.json
+model:
+  name: efficientnet_unet
+  encoder: efficientnet-b4
+  encoder_weights: imagenet
+  attention: scse
+  in_channels: 3
+  classes: 2
+  tile_size: 768
+training:
+  epochs: 120
+  batch_size: 2
+  learning_rate: 4.0e-05
+  optimizer: adamw
+  scheduler: cosine
+  loss: boundary_head
+  augmentation: true
+  num_workers: 2
+  n_folds: 5
+  fold_index: 3
+  early_stopping_patience: 30
+  early_stopping_min_delta: 0.0001
+  init_from: runs/civilized-wars-52/checkpoint.pth
+loss:
+  boundary_weight: 0.5
+  bce_weight: 0.5
+  lovasz_weight: 0.5
+wandb:
+  project: block-vectorization
+  entity: research_carto

worthy-blaze-63/checkpoint.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1db78df8daef4dbdc54e9aed65da7b168e5937addc580afc21552a54088b2a4
+size 82028033

worthy-blaze-63/config.yaml

@@ -0,0 +1,33 @@
+experiment_name: cv_fold4
+preprocessing:
+  approach: v2
+  manifest_path: data/preprocessed/v2_manifest_cv.json
+model:
+  name: efficientnet_unet
+  encoder: efficientnet-b4
+  encoder_weights: imagenet
+  attention: scse
+  in_channels: 3
+  classes: 2
+  tile_size: 768
+training:
+  epochs: 120
+  batch_size: 2
+  learning_rate: 4.0e-05
+  optimizer: adamw
+  scheduler: cosine
+  loss: boundary_head
+  augmentation: true
+  num_workers: 2
+  n_folds: 5
+  fold_index: 4
+  early_stopping_patience: 30
+  early_stopping_min_delta: 0.0001
+  init_from: runs/civilized-wars-52/checkpoint.pth
+loss:
+  boundary_weight: 0.5
+  bce_weight: 0.5
+  lovasz_weight: 0.5
+wandb:
+  project: block-vectorization
+  entity: research_carto