File size: 29,554 Bytes
ca87569 |
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 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 |
{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"private_outputs": true,
"provenance": [],
"machine_shape": "hm"
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
}
},
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "CSC6_ShCp6h9"
},
"outputs": [],
"source": [
"!unzip AI.zip\n",
"!unzip Photo.zip"
]
},
{
"cell_type": "code",
"source": [
"!pip install umap-learn\n",
"!pip install PyWavelets"
],
"metadata": {
"id": "N6CWTCziLMbf"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import accuracy_score, f1_score, confusion_matrix, ConfusionMatrixDisplay\n",
"from sklearn.preprocessing import StandardScaler\n",
"from sklearn.decomposition import PCA\n",
"import umap\n",
"import pywt"
],
"metadata": {
"id": "53ZvG8NbATlR"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Create a function to load all the files in a folder as images.\n",
"\n",
"import os\n",
"from PIL import Image\n",
"def load_images_from_folder(folder):\n",
" images = []\n",
" labels = []\n",
" for filename in os.listdir(folder):\n",
" if not filename.endswith('.jpg') and not filename.endswith('.png') \\\n",
" and not filename.endswith('jpeg') and not filename.endswith('webp'):\n",
" continue\n",
" img = Image.open(os.path.join(folder,filename))\n",
" img = img.resize((512, 512))\n",
" if img is not None:\n",
" images.append(img)\n",
" labels.append(1 if folder == \"AI\" else 0)\n",
" return images, labels"
],
"metadata": {
"id": "BH6bOWUXsi_D"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you write a function that can implement the discrete wavelet transform and display the wavelets given in an array for the image? The function should take in an image_path and a list of wavelets and perform the dwt and display the wavelets.\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"def apply_wavelet_transform_and_display_multiple(image_path, wavelets):\n",
" # Load the image\n",
" img = Image.open(image_path).convert('L')\n",
"\n",
" # Convert image to numpy array\n",
" img_array = np.array(img)\n",
"\n",
" num_wavelets = len(wavelets)\n",
" fig, axes = plt.subplots(1, num_wavelets + 1, figsize=(5 * (num_wavelets + 1), 5))\n",
"\n",
" # Display the original image\n",
" axes[0].imshow(img_array, cmap='gray')\n",
" axes[0].set_title('Original Image')\n",
"\n",
" # Apply DWT and display wavelets\n",
" for i, wavelet in enumerate(wavelets):\n",
" cA, cD = pywt.dwt(img_array, wavelet)\n",
" axes[i + 1].imshow(cD, cmap='gray')\n",
" axes[i + 1].set_title(f'Approximate Image ({wavelet})')\n",
"\n",
" plt.tight_layout()\n",
" plt.show()\n"
],
"metadata": {
"id": "sBRFYk0C2nfX"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"apply_wavelet_transform_and_display_multiple('kiri-in-high-resolution-love-her-3-v0-ezejx6try3va1.webp', ['db1', 'db6', 'db10', 'db12', 'db16'])"
],
"metadata": {
"id": "KfY3qSfkxJnS"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you write a function that given a list of images from PIL can convert them to grayscale and apply a set of wavelets using dwt and then combined them into one feature vector?\n",
"\n",
"import numpy as np\n",
"def extract_wavelet_features(images, wavelets):\n",
" all_features = []\n",
" for img in images:\n",
" img_gray = img.convert('L')\n",
" img_array = np.array(img_gray)\n",
" features = []\n",
" for wavelet in wavelets:\n",
" cA, cD = pywt.dwt(img_array, wavelet)\n",
" features.extend(cD.flatten())\n",
" all_features.append(features)\n",
" return np.array(all_features)\n"
],
"metadata": {
"id": "ufMhM7_86IbC"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Apply the Fourier transform to the images from the load_images_from_folder function.\n",
"\n",
"import numpy as np\n",
"\n",
"\n",
"# Example usage (assuming 'folder_path' contains your images)\n",
"ai_images, ai_labels = load_images_from_folder('AI')\n",
"photo_images, photo_labels = load_images_from_folder('Photo')\n",
"min_length = min(len(ai_images), len(photo_images))\n",
"ai_images = ai_images[:min_length]\n",
"photo_images = photo_images[:min_length]\n",
"ai_labels = ai_labels[:min_length]\n",
"photo_labels = photo_labels[:min_length]\n",
"\n",
"print(f\"Number of AI images: {len(ai_images)}\")\n",
"print(f\"Number of Photo images: {len(photo_images)}\")\n",
"images = ai_images + photo_images\n",
"labels = ai_labels + photo_labels\n",
"features = np.array(extract_wavelet_features(images, [\"db4\", \"db10\"]))"
],
"metadata": {
"id": "7Pfn_0-QswSh"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"reducer = umap.UMAP(n_neighbors=16, n_components=32, random_state=42)\n",
"embeddings = reducer.fit_transform(features)"
],
"metadata": {
"id": "xc_1hAuTLdUj"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"reducer.embedding_.dtype"
],
"metadata": {
"id": "qprQSJTCaPpv"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"X_train, X_test, y_train, y_test = train_test_split(embeddings, labels, test_size=0.2, random_state=42)"
],
"metadata": {
"id": "dFQYuL3MbJLj"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from xgboost import XGBClassifier"
],
"metadata": {
"id": "HoySyJJ4cL3n"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"xgb_clf = XGBClassifier(n_estimators=200, eval_metric=\"logloss\", learning_rate=0.01,\n",
" reg_lambda=0.8, max_depth=5, gamma=1.0, subsample=0.5,\n",
" colsample_bytree=0.5, min_child_weight=10)\n",
"xgb_clf.fit(X_train, y_train, eval_set=[(X_test, y_test)],\n",
" verbose=True)\n",
"\n",
"xgb_clf_pred = xgb_clf.predict(X_test)\n",
"score = xgb_clf.score(X_test, y_test)\n",
"print(f\"Accuracy: {score}\")\n",
"\n",
"print(f\"F1 score: {f1_score(y_test, xgb_clf_pred)}\")"
],
"metadata": {
"id": "vP5jesFXJHcY"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Calculate the training accuracy\n",
"\n",
"xgb_clf_pred_train = xgb_clf.predict(X_train)\n",
"score = xgb_clf.score(X_train, y_train)\n",
"print(f\"Training Accuracy: {score}\")\n",
"\n",
"score = xgb_clf.score(X_test, y_test)\n",
"print(f\"Test Accuracy: {score}\")"
],
"metadata": {
"id": "IljcJVxVVlgI"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you perform four fold cross validation on the xgboost model?\n",
"\n",
"from sklearn.model_selection import cross_val_score, KFold\n",
"# Perform four-fold cross-validation\n",
"kfold = KFold(n_splits=4, shuffle=True, random_state=42)\n",
"scores = cross_val_score(xgb_clf, embeddings, labels, cv=kfold, scoring='accuracy')\n",
"\n",
"# Print the cross-validation scores\n",
"print(\"Cross-validation scores:\", scores)\n",
"print(\"Average cross-validation score:\", scores.mean())"
],
"metadata": {
"id": "peofLwk78-mE"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"ConfusionMatrixDisplay.from_estimator(xgb_clf, X_test, y_test)"
],
"metadata": {
"id": "5GvVgOoXcbJ-"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"xgb_clf.save_model(\"xgb_flux_detection_model.json\")"
],
"metadata": {
"id": "5TZsByCxQqbU"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: A random classifier\n",
"\n",
"from sklearn.dummy import DummyClassifier\n",
"\n",
"# Initialize a random classifier\n",
"dummy_clf = DummyClassifier(strategy='uniform') # Predicts randomly\n",
"\n",
"# Fit the classifier (not really necessary for a random classifier)\n",
"dummy_clf.fit(X_train, y_train)\n",
"\n",
"# Make predictions\n",
"dummy_pred = dummy_clf.predict(X_test)\n",
"\n",
"# Evaluate the performance\n",
"score = dummy_clf.score(X_test, y_test)\n",
"print(f\"Accuracy: {score}\")\n",
"print(f\"F1 score: {f1_score(y_test, dummy_pred)}\")\n",
"\n",
"ConfusionMatrixDisplay.from_estimator(dummy_clf, X_test, y_test)"
],
"metadata": {
"id": "X7qkISlS4QjW"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: random forests with pruning\n",
"\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"\n",
"# Initialize the RandomForestClassifier with pruning parameters\n",
"rf_clf = RandomForestClassifier(n_estimators=100, # Number of trees in the forest\n",
" max_depth=5, # Maximum depth of each tree (pruning)\n",
" min_samples_split=5, # Minimum samples required to split a node (pruning)\n",
" random_state=42) # Random seed for reproducibility\n",
"\n",
"# Fit the classifier to the training data\n",
"rf_clf.fit(X_train, y_train)\n",
"\n",
"# Make predictions on the test data\n",
"rf_pred = rf_clf.predict(X_test)\n",
"\n",
"# Evaluate the performance\n",
"score = rf_clf.score(X_test, y_test)\n",
"print(f\"Accuracy: {score}\")\n",
"\n",
"print(f\"F1 score: {f1_score(y_test, rf_pred)}\")\n",
"\n",
"ConfusionMatrixDisplay.from_estimator(rf_clf, X_test, y_test)"
],
"metadata": {
"id": "3qJFLsYT3xmi"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you perform four fold cross validation on the rf model?\n",
"\n",
"from sklearn.model_selection import cross_val_score, KFold\n",
"# Perform four-fold cross-validation\n",
"kfold = KFold(n_splits=4, shuffle=True, random_state=42)\n",
"scores = cross_val_score(rf_clf, embeddings, labels, cv=kfold, scoring='accuracy')\n",
"\n",
"# Print the cross-validation scores\n",
"print(\"Cross-validation scores:\", scores)\n",
"print(\"Average cross-validation score:\", scores.mean())"
],
"metadata": {
"id": "-gDc0KvD9_Yp"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: SVC classifier\n",
"\n",
"from sklearn.svm import SVC\n",
"\n",
"# Initialize the SVC classifier\n",
"svc_clf = SVC()\n",
"\n",
"# Fit the classifier to the training data\n",
"svc_clf.fit(X_train, y_train)\n",
"\n",
"# Make predictions on the test data\n",
"svc_pred = svc_clf.predict(X_test)\n",
"\n",
"# Evaluate the performance\n",
"score = svc_clf.score(X_test, y_test)\n",
"print(f\"Accuracy: {score}\")\n",
"\n",
"print(f\"F1 score: {f1_score(y_test, svc_pred)}\")\n",
"\n",
"ConfusionMatrixDisplay.from_estimator(svc_clf, X_test, y_test)\n"
],
"metadata": {
"id": "1sQjrGeZ8Ir3"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: classify with KNN and K=7\n",
"\n",
"from sklearn.neighbors import KNeighborsClassifier\n",
"# Initialize the KNeighborsClassifier with K=7\n",
"knn_clf = KNeighborsClassifier(n_neighbors=7)\n",
"\n",
"# Fit the classifier to the training data\n",
"knn_clf.fit(X_train, y_train)\n",
"\n",
"# Make predictions on the test data\n",
"knn_pred = knn_clf.predict(X_test)\n",
"\n",
"# Evaluate the performance\n",
"score = knn_clf.score(X_test, y_test)\n",
"print(f\"Accuracy: {score}\")\n",
"\n",
"print(f\"F1 score: {f1_score(y_test, knn_pred)}\")\n",
"\n",
"ConfusionMatrixDisplay.from_estimator(knn_clf, X_test, y_test)\n"
],
"metadata": {
"id": "vU8SRYsZ72Sr"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you perform four fold cross validation on the KNN model?\n",
"\n",
"from sklearn.model_selection import cross_val_score, KFold\n",
"# Perform four-fold cross-validation\n",
"kfold = KFold(n_splits=4, shuffle=True, random_state=42)\n",
"scores = cross_val_score(knn_clf, embeddings, labels, cv=kfold, scoring='accuracy')\n",
"\n",
"# Print the cross-validation scores\n",
"print(\"Cross-validation scores:\", scores)\n",
"print(\"Average cross-validation score:\", scores.mean())"
],
"metadata": {
"id": "1X9_4kAKRlSm"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"import plotly.express as px\n",
"# Initialize UMAP with desired parameters\n",
"reducer = umap.UMAP(n_components=2, random_state=42)\n",
"\n",
"# Reduce the dimensionality of the features array\n",
"embedding = reducer.fit_transform(features)\n",
"import pandas as pd\n",
"\n",
"# Create a DataFrame for Plotly\n",
"embedding_df = pd.DataFrame(embedding, columns=['UMAP1', 'UMAP2'])\n",
"embedding_df['label'] = labels\n",
"# Create a scatter plot\n",
"fig = px.scatter(\n",
" embedding_df,\n",
" x='UMAP1',\n",
" y='UMAP2',\n",
" color='label',\n",
" title='UMAP Dimensionality Reduction',\n",
" labels={'color': 'Label'}\n",
")\n",
"\n",
"# Show the plot\n",
"fig.show()"
],
"metadata": {
"id": "wMEQoDF2Goj-"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Save the knn classifier as a file\n",
"\n",
"import joblib\n",
"\n",
"# Save the knn classifier to a file\n",
"filename = 'knn_model.pkl'\n",
"joblib.dump(knn_clf, filename)\n"
],
"metadata": {
"id": "I-Myacr4zsVy"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: load the knn model\n",
"\n",
"# Load the knn classifier from the file\n",
"filename = 'knn_model.pkl'\n",
"loaded_knn_clf = joblib.load(filename)"
],
"metadata": {
"id": "yayMkQELAbZO"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: load the validation images and apply the wavelet transforms\n",
"\n",
"# Assuming 'validation_folder' contains your validation images\n",
"validation_images, validation_labels = load_images_from_folder('validation_folder')\n",
"\n",
"# Extract wavelet features from validation images\n",
"validation_features = extract_wavelet_features(validation_images, [\"db4\", \"db10\"])\n",
"\n",
"# Reduce dimensionality of validation features using the same UMAP reducer\n",
"validation_embeddings = reducer.transform(validation_features)\n",
"\n",
"# Now you have 'validation_embeddings' and 'validation_labels' for further use\n",
"# (e.g., evaluating your trained models on validation data)\n"
],
"metadata": {
"id": "GKCz35S8E9jn"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"### Validation"
],
"metadata": {
"id": "nrcTRu_ilEGk"
}
},
{
"cell_type": "code",
"source": [
"!unzip Validation.zip"
],
"metadata": {
"id": "Yajcb-E5lDgl"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: load the validation images\n",
"\n",
"# Assuming 'Validation' is the folder containing your validation images\n",
"ai_validation_images, ai_validation_labels = load_images_from_folder('Validation/AI')\n",
"photo_validation_images, photo_validation_labels = load_images_from_folder('Validation/Photo')\n",
"\n",
"\n",
"# Now you have 'validation_images' and 'validation_labels' for further use\n",
"print(f\"Number of AI Validation images: {len(ai_validation_images)}\")\n",
"print(f\"Number of Photo Validation images: {len(ai_validation_images)}\")"
],
"metadata": {
"id": "mS8hzT-TlGER"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Combine both validation datasets and extract the wavelet features.\n",
"\n",
"# Combine validation datasets\n",
"validation_images = ai_validation_images + photo_validation_images\n",
"validation_labels = ai_validation_labels + photo_validation_labels\n",
"\n",
"# Extract wavelet features from validation images\n",
"validation_features = extract_wavelet_features(validation_images, [\"db4\", \"db10\"])"
],
"metadata": {
"id": "iTeZUqEblbu1"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: apply the reducer to find the validation embeddings\n",
"\n",
"# Reduce dimensionality of validation features using the same UMAP reducer\n",
"validation_embeddings = reducer.transform(validation_features)"
],
"metadata": {
"id": "jdUbmE4Hltng"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: find the accuracy and f1 score on the knn classifier for validation features\n",
"\n",
"# Make predictions on the validation data\n",
"knn_pred_validation = knn_clf.predict(validation_embeddings)\n",
"\n",
"# Evaluate the performance on validation data\n",
"score_validation = knn_clf.score(validation_embeddings, validation_labels)\n",
"print(f\"Validation Accuracy: {score_validation}\")\n",
"\n",
"print(f\"Validation F1 score: {f1_score(validation_labels, knn_pred_validation)}\")\n"
],
"metadata": {
"id": "ls2ij5VxlyOX"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: Can you combine the entire pipeline into one class?\n",
"\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import accuracy_score, f1_score, confusion_matrix, ConfusionMatrixDisplay\n",
"from sklearn.preprocessing import StandardScaler\n",
"from sklearn.decomposition import PCA\n",
"import umap\n",
"import pywt\n",
"import os\n",
"from PIL import Image\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from xgboost import XGBClassifier\n",
"from sklearn.model_selection import cross_val_score, KFold\n",
"from sklearn.dummy import DummyClassifier\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.svm import SVC\n",
"from sklearn.neighbors import KNeighborsClassifier\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import classification_report\n",
"import plotly.express as px\n",
"import pandas as pd\n",
"import joblib\n",
"from tqdm import tqdm\n",
"import lzma\n",
"\n",
"class FluxClassifier:\n",
" def __init__(self, wavelets=[\"db4\", \"db10\"], umap_n_neighbors=16, umap_n_components=32, random_state=42):\n",
" self.wavelets = wavelets\n",
" self.umap_n_neighbors = umap_n_neighbors\n",
" self.umap_n_components = umap_n_components\n",
" self.random_state = random_state\n",
" self.reducer = umap.UMAP(n_neighbors=self.umap_n_neighbors,\n",
" n_components=self.umap_n_components,\n",
" random_state=self.random_state)\n",
" self.classifier = KNeighborsClassifier(n_neighbors=7) # Default classifier\n",
"\n",
" def load_images_from_folder(self, folder):\n",
" images = []\n",
" labels = []\n",
" print(f\"Loading images from {folder}\")\n",
" for filename in tqdm(os.listdir(folder)):\n",
" if not (filename.endswith('.jpg') or filename.endswith('.png') or\n",
" filename.endswith('jpeg') or filename.endswith('webp')):\n",
" continue\n",
" img = Image.open(os.path.join(folder, filename))\n",
" img = img.resize((512, 512))\n",
" if img is not None:\n",
" images.append(img)\n",
" labels.append(1 if \"AI\" in folder else 0) # Assuming folder names contain \"AI\" or not\n",
" return images, labels\n",
"\n",
" def extract_wavelet_features(self, images):\n",
" all_features = []\n",
" for img in images:\n",
" img_gray = img.convert('L')\n",
" img_array = np.array(img_gray)\n",
" features = []\n",
" for wavelet in self.wavelets:\n",
" cA, cD = pywt.dwt(img_array, wavelet)\n",
" features.extend(cD.flatten())\n",
" all_features.append(features)\n",
" return np.array(all_features)\n",
"\n",
" def fit(self, train_folder1, train_folder2):\n",
" # Load images and extract features\n",
" images1, labels1 = self.load_images_from_folder(train_folder1)\n",
" images2, labels2 = self.load_images_from_folder(train_folder2)\n",
"\n",
" min_length = min(len(images1), len(images2))\n",
" images1 = images1[:min_length]\n",
" images2 = images2[:min_length]\n",
" labels1 = labels1[:min_length]\n",
" labels2 = labels2[:min_length]\n",
"\n",
" images = images1 + images2\n",
" labels = labels1 + labels2\n",
" features = self.extract_wavelet_features(images)\n",
"\n",
" # Apply UMAP dimensionality reduction\n",
" embeddings = self.reducer.fit_transform(features)\n",
" X_train, X_test, y_train, y_test = train_test_split(embeddings, labels, test_size=0.2, random_state=42)\n",
"\n",
" # Train the classifier\n",
" self.classifier.fit(X_train, y_train)\n",
"\n",
" acc = self.classifier.score(X_test, y_test)\n",
" y_pred = self.classifier.predict(X_test)\n",
" print(f\"Classifier accuracy = {acc}\")\n",
"\n",
" f1 = f1_score(y_test, y_pred)\n",
" print(f\"Classifier F1 = {f1}\")\n",
" print(classification_report(y_test, y_pred))\n",
"\n",
"\n",
" def predict(self, images):\n",
" # Load images and extract features\n",
" features = self.extract_wavelet_features(images)\n",
"\n",
" # Apply UMAP dimensionality reduction\n",
" embeddings = self.reducer.transform(features)\n",
"\n",
" # Make predictions\n",
" return self.classifier.predict(embeddings)\n",
"\n",
" def predict_proba(self, images):\n",
" # Load images and extract features\n",
" features = self.extract_wavelet_features(images)\n",
"\n",
" # Apply UMAP dimensionality reduction\n",
" embeddings = self.reducer.transform(features)\n",
"\n",
" # Make predictions\n",
" return self.classifier.predict_proba(embeddings)\n",
"\n",
" def score(self, test_folder):\n",
" # Load images and extract features\n",
" images, labels = self.load_images_from_folder(test_folder)\n",
" features = self.extract_wavelet_features(images)\n",
"\n",
" # Apply UMAP dimensionality reduction\n",
" embeddings = self.reducer.transform(features)\n",
"\n",
" # Evaluate the classifier\n",
" return self.classifier.score(embeddings, labels)\n",
"\n",
" def save_model(self, filename):\n",
" joblib.dump(self, filename, compress=('zlib', 9))\n",
"\n",
" @staticmethod\n",
" def load_model(filename):\n",
" return joblib.load(filename)"
],
"metadata": {
"id": "V8NO_N4QteQK"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"classifier = FluxClassifier()\n",
"classifier.fit(\"AI\", \"Photo\")"
],
"metadata": {
"id": "sFYjKz1L6xgg"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"classifier.save_model(\"flux_classifier.pkl\")"
],
"metadata": {
"id": "tiLVrOTF_ZGM"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# prompt: save the model to my google drive.\n",
"\n",
"from google.colab import drive\n",
"drive.mount('/content/drive')\n",
"!cp flux_classifier.pkl /content/drive/MyDrive"
],
"metadata": {
"id": "sXo1mHFSADuS"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"images = [Image.open(\"pDGQUK1BYaJYhrFB5ouQU.jpeg\"), Image.open(\"jenta2.jpeg\")]\n",
"predictions = classifier.predict_proba(images)\n",
"print(predictions)"
],
"metadata": {
"id": "cNVwQ7Oq6vWa"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [],
"metadata": {
"id": "98TbK3uH-_CD"
},
"execution_count": null,
"outputs": []
}
]
} |