File size: 2,121 Bytes
16a0f31 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 |
from tqdm import tqdm
import os
from pathlib import Path
import numpy as np
from evaluation.utils.json_helpers import dict_to_json
# -----------------------------------------------------------------------------
# This class is a sample code provided to help you with saving your predicted
# scores as JSON files. We strongly suggest that you use the provided methods,
# but you are NOT required to follow this structure. Feel free to adapt,
# modify, or extend this template in favor of your own workflow.
# -----------------------------------------------------------------------------
class DumpScores:
def __init__(self):
self.scores_dir = "./output_scores"
self.save_scores_precision = 4
def save_scores(self, image_path_list, pred_img_level, pred_pix_level):
print(
f"Saving scores at '{self.scores_dir}' with precision: '{self.save_scores_precision}'"
)
for i in tqdm(range(len(image_path_list)), desc=f"Saving scores"):
image_path = image_path_list[i]
image_score_path = self.get_scores_path_for_image(image_path)
os.makedirs(os.path.dirname(image_score_path), exist_ok=True)
vectorized_enforce_precision = np.vectorize(self.enforce_precision)
d = {
"img_level_score": vectorized_enforce_precision(
pred_img_level[i], self.save_scores_precision
),
"pix_level_score": vectorized_enforce_precision(
pred_pix_level[i], self.save_scores_precision
),
}
dict_to_json(d, image_score_path)
def get_scores_path_for_image(self, image_path):
"""example image_path: './data/photovoltaic_module/test/good/037.png'"""
path = Path(image_path)
category, split, anomaly_type = path.parts[-4:-1]
image_name = path.stem
return os.path.join(
self.scores_dir, category, split, anomaly_type, f"{image_name}_scores.json"
)
def enforce_precision(self, x, precision):
return float(f"{x:.{precision}f}")
|