| """Re-score a Blue Marble Flood model on Sen1Floods11: the official hand-labelled test split and the held-out Bolivia split. |
| |
| pip install numpy onnxruntime rasterio huggingface_hub |
| python evaluate.py --model nano/model_fp32.onnx # or nano/model_int8.onnx |
| |
| Data: Sen1Floods11 v1.1 (hand-labelled Sentinel-2 L1C chips and labels, official splits), about 1.8 GB, downloaded once.""" |
| import numpy as np |
|
|
| |
| MEAN = np.array([2357.089, 2137.385, 2018.788, 2082.986, 2295.651, 2854.537, 3122.849, 3040.560, 3306.481, 1473.847, 506.070, 2472.825, 1838.929], np.float32) |
| STD = np.array([1624.683, 1675.806, 1557.708, 1833.702, 1823.738, 1733.977, 1732.131, 1679.732, 1727.26, 1024.687, 442.165, 1331.411, 1160.419], np.float32) |
|
|
|
|
| def normalize(dn, bands=None): |
| """dn: [C, H, W] digital numbers for the 13 bands, or for the band indices in `bands`. Missing bands are 0 after normalisation.""" |
| out = np.zeros((13, *dn.shape[1:]), np.float32) |
| idx = list(range(13)) if bands is None else list(bands) |
| out[idx] = (dn.astype(np.float32) - MEAN[idx, None, None]) / STD[idx, None, None] |
| return out |
|
|
|
|
| def session(path, threads=0): |
| import onnxruntime as ort |
| so = ort.SessionOptions() |
| if threads: so.intra_op_num_threads = threads |
| prov = [p for p in ("CUDAExecutionProvider", "CPUExecutionProvider") if p in ort.get_available_providers()] |
| return ort.InferenceSession(path, so, providers=prov) |
|
|
|
|
| def predict(sess, x): |
| """Whole image in one pass: pad H and W to multiples of 16 by edge replication, crop the logits back.""" |
| h, w = x.shape[-2:] |
| xp = np.pad(x, ((0, 0), (0, (-h) % 16), (0, (-w) % 16)), mode="edge")[None].astype(np.float32) |
| return sess.run(None, {sess.get_inputs()[0].name: xp})[0][0, :, :h, :w].argmax(0) |
|
|
|
|
| class MIoU: |
| def __init__(self, classes): |
| self.classes, n = classes, len(classes) |
| self.cm = np.zeros((n, n), np.int64) |
|
|
| def add(self, y, p): |
| n = len(self.classes); m = y >= 0 |
| self.cm += np.bincount(n * y[m].astype(np.int64) + p[m], minlength=n * n).reshape(n, n) |
|
|
| def report(self, title): |
| inter = np.diag(self.cm).astype(float); union = self.cm.sum(0) + self.cm.sum(1) - inter |
| iou = inter / np.maximum(union, 1) |
| print(f"{title}: mIoU {100 * iou.mean():.2f} (" + ", ".join(f"{c} {100 * v:.2f}" for c, v in zip(self.classes, iou)) + ")", flush=True) |
| return float(iou.mean()) |
|
|
|
|
| import argparse, os, tarfile |
| import numpy as np, rasterio |
|
|
| SPLITS = {"test": "flood_test_data", "bolivia": "flood_bolivia_data"} |
|
|
|
|
| def fetch(data): |
| root = f"{data}/sen1floods11_v1.1" |
| if os.path.exists(f"{root}/.ok"): return root |
| from huggingface_hub import hf_hub_download |
| arc = f"{data}/sen1floods11_v1_1.tar.gz" |
| if not os.path.exists(arc): arc = hf_hub_download("blumenstiel/Sen1Floods11", "sen1floods11_v1_1.tar.gz", repo_type="dataset", local_dir=data) |
| with tarfile.open(arc, "r|gz") as t: |
| for m in t: |
| if "/splits/" in m.name and m.isfile(): t.extract(m, data, filter="data") |
| ids = {i.strip() for s in SPLITS.values() for i in open(f"{root}/splits/{s}.txt") if i.strip()} |
| want = {f"{i}_S2Hand.tif" for i in ids} | {f"{i}_LabelHand.tif" for i in ids} |
| with tarfile.open(arc, "r|gz") as t: |
| for m in t: |
| if os.path.basename(m.name) in want and ("/S2L1CHand/" in m.name or "/LabelHand/" in m.name): t.extract(m, data, filter="data") |
| open(f"{root}/.ok", "w").close() |
| return root |
|
|
|
|
| def read(p): |
| with rasterio.open(p) as s: return s.read() |
|
|
|
|
| def main(): |
| ap = argparse.ArgumentParser(); ap.add_argument("--model", required=True); ap.add_argument("--data", default="data") |
| a = ap.parse_args(); os.makedirs(a.data, exist_ok=True) |
| root = fetch(a.data); sess = session(a.model) |
| for split, f in SPLITS.items(): |
| ids = [l.strip() for l in open(f"{root}/splits/{f}.txt") if l.strip()] |
| m = MIoU(["not water", "water"]) |
| for i in ids: |
| dn = np.clip(np.nan_to_num(read(f"{root}/data/S2L1CHand/{i}_S2Hand.tif")).astype(np.int32), 0, 65535).astype(np.uint16) |
| y = read(f"{root}/data/LabelHand/{i}_LabelHand.tif")[0].astype(np.int64) |
| m.add(y, predict(sess, normalize(dn))) |
| m.report(f"Sen1Floods11 {split} ({len(ids)} chips)") |
|
|
|
|
| if __name__ == "__main__": main() |
|
|
