Download train_svm.py from siyah1/Fdet: direct link, hf CLI and curl.
- Browser
- Download file 1.34 kB
-
https://huggingface.co/spaces/siyah1/Fdet/resolve/main/train_svm.py
- Command line
-
hf download hf://spaces/siyah1/Fdet/train_svm.py
-
curl -L -o train_svm.py https://huggingface.co/spaces/siyah1/Fdet/resolve/main/train_svm.py
1.34 kB
| import os | |
| import cv2 | |
| import numpy as np | |
| import insightface | |
| from sklearn.svm import SVC | |
| import joblib | |
| from tqdm import tqdm | |
| # 1. ArcFace ๋ชจ๋ธ ๋ก๋ฉ | |
| model = insightface.app.FaceAnalysis(name="buffalo_l", providers=['CPUExecutionProvider']) | |
| model.prepare(ctx_id=0) | |
| X = [] | |
| y = [] | |
| # 2. ์ผ๊ตด ์๋ฒ ๋ฉ ์ถ์ถ ํจ์ | |
| def extract_embeddings(folder_path, label): | |
| for fname in tqdm(os.listdir(folder_path), desc=f"{label} - {os.path.basename(folder_path)}"): | |
| if not fname.lower().endswith((".jpg", ".jpeg", ".png")): | |
| continue | |
| path = os.path.join(folder_path, fname) | |
| img = cv2.imread(path) | |
| if img is None: | |
| continue | |
| faces = model.get(img) | |
| if faces: | |
| emb = faces[0].embedding | |
| X.append(emb) | |
| y.append(label) | |
| # 3. ์ฌ๋(1) / ๋น์ฌ๋(0) ํด๋๋ก๋ถํฐ ์๋ฒ ๋ฉ ์ถ์ถ | |
| extract_embeddings("dataset/human", 1) | |
| extract_embeddings("dataset/nonhuman", 0) | |
| print(f"\nโ ์ด ์ํ ์: {len(X)}") | |
| print(f" - ์ฌ๋ ์ผ๊ตด: {y.count(1)}") | |
| print(f" - ๋น์ฌ๋: {y.count(0)}") | |
| # 4. SVM ๋ถ๋ฅ๊ธฐ ํ์ต | |
| clf = SVC(kernel='linear', probability=True) | |
| clf.fit(X, y) | |
| # 5. ๋ชจ๋ธ ์ ์ฅ | |
| joblib.dump(clf, "is_human_classifier.pkl") | |
| print("\nโ ๋ถ๋ฅ๊ธฐ ์ ์ฅ ์๋ฃ: is_human_classifier.pkl") | |