{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Computer Vision: Clasificación de Imágenes (pre-trained)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'c:\\\\Users\\\\CynYDie\\\\Desktop\\\\UTN_Haedo\\\\Clases\\\\Clase14_CV'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "os.getcwd()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Clasificar imagen con ResNet50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 4s/step\n",
      "Predicted:\n",
      "n02124075 (Egyptian_cat): 0.38\n",
      "n02123159 (tiger_cat): 0.16\n",
      "n02123045 (tabby): 0.16\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.applications import ResNet50\n",
    "from tensorflow.keras.applications.resnet50 import preprocess_input, decode_predictions\n",
    "import numpy as np\n",
    "\n",
    "resnet50_model = ResNet50(weights='imagenet')\n",
    "\n",
    "img_path = 'datasets/cat.jpeg'\n",
    "try:\n",
    "    img = tf.keras.utils.load_img(img_path, target_size=(224, 224))\n",
    "except FileNotFoundError:\n",
    "    print(f\"Error: Image file not found at {img_path}\")\n",
    "    exit(1)\n",
    "\n",
    "x = tf.keras.utils.img_to_array(img)\n",
    "x = np.expand_dims(x, axis=0)\n",
    "x = preprocess_input(x)\n",
    "\n",
    "preds = resnet50_model.predict(x)\n",
    "print('Predicted:')\n",
    "for label, class_id, confidence in decode_predictions(preds, top=3)[0]:\n",
    "    print(f\"{label} ({class_id}): {confidence:.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Clasificar imagen con VGG16"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 4s/step\n",
      "Predicted:\n",
      "n02123159 (tiger_cat): 0.29\n",
      "n02123045 (tabby): 0.25\n",
      "n02124075 (Egyptian_cat): 0.24\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.applications import VGG16\n",
    "from tensorflow.keras.applications.vgg16 import preprocess_input, decode_predictions\n",
    "import numpy as np\n",
    "\n",
    "vgg16_model = VGG16(weights='imagenet')\n",
    "\n",
    "img_path = 'datasets/cat.jpeg'\n",
    "try:\n",
    "    img = tf.keras.utils.load_img(img_path, target_size=(224, 224))\n",
    "except FileNotFoundError:\n",
    "    print(f\"Error: Image file not found at {img_path}\")\n",
    "    exit(1)\n",
    "\n",
    "x = tf.keras.utils.img_to_array(img)\n",
    "x = np.expand_dims(x, axis=0)\n",
    "x = preprocess_input(x)\n",
    "\n",
    "preds = vgg16_model.predict(x)\n",
    "print('Predicted:')\n",
    "for label, class_id, confidence in decode_predictions(preds, top=3)[0]:\n",
    "    print(f\"{label} ({class_id}): {confidence:.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Clasificar imagen con distintos modelos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1s/step\n",
      "WARNING:tensorflow:5 out of the last 5 calls to <function TensorFlowTrainer.make_predict_function.<locals>.one_step_on_data_distributed at 0x0000029547C85580> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 3s/step\n",
      "WARNING:tensorflow:6 out of the last 6 calls to <function TensorFlowTrainer.make_predict_function.<locals>.one_step_on_data_distributed at 0x0000029547C86DE0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2s/step\n",
      "VGG16:\n",
      "n02123159 (tiger_cat): 0.29\n",
      "n02123045 (tabby): 0.25\n",
      "n02124075 (Egyptian_cat): 0.24\n",
      "\n",
      "ResNet50:\n",
      "n02124075 (Egyptian_cat): 0.38\n",
      "n02123159 (tiger_cat): 0.16\n",
      "n02123045 (tabby): 0.16\n",
      "\n",
      "MobileNetV2:\n",
      "n02123159 (tiger_cat): 0.64\n",
      "n02123045 (tabby): 0.13\n",
      "n02124075 (Egyptian_cat): 0.02\n",
      "\n",
      "Comparison of top 3 results:\n",
      "Rank 1: VGG16=n02123159 (tiger_cat), ResNet50=n02124075 (Egyptian_cat), MobileNetV2=n02123159 (tiger_cat)\n",
      "Rank 2: VGG16=n02123045 (tabby), ResNet50=n02123159 (tiger_cat), MobileNetV2=n02123045 (tabby)\n",
      "Rank 3: VGG16=n02124075 (Egyptian_cat), ResNet50=n02123045 (tabby), MobileNetV2=n02124075 (Egyptian_cat)\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.applications import VGG16, ResNet50, MobileNetV2\n",
    "from tensorflow.keras.applications.vgg16 import preprocess_input as vgg16_preprocess\n",
    "from tensorflow.keras.applications.resnet50 import preprocess_input as resnet50_preprocess\n",
    "from tensorflow.keras.applications.mobilenet_v2 import preprocess_input as mobilenetv2_preprocess\n",
    "from tensorflow.keras.applications.vgg16 import decode_predictions\n",
    "import numpy as np\n",
    "\n",
    "# Load pre-trained models\n",
    "vgg16_model = VGG16(weights='imagenet')\n",
    "resnet50_model = ResNet50(weights='imagenet')\n",
    "mobilenetv2_model = MobileNetV2(weights='imagenet')\n",
    "\n",
    "# Load image\n",
    "img_path = 'datasets/cat.jpeg'\n",
    "try:\n",
    "    img = tf.keras.utils.load_img(img_path, target_size=(224, 224))\n",
    "except FileNotFoundError:\n",
    "    print(f\"Error: Image file not found at {img_path}\")\n",
    "    exit(1)\n",
    "\n",
    "# Preprocess image for each model\n",
    "vgg16_x = tf.keras.utils.img_to_array(img)\n",
    "vgg16_x = np.expand_dims(vgg16_x, axis=0)\n",
    "vgg16_x = vgg16_preprocess(vgg16_x)\n",
    "\n",
    "resnet50_x = tf.keras.utils.img_to_array(img)\n",
    "resnet50_x = np.expand_dims(resnet50_x, axis=0)\n",
    "resnet50_x = resnet50_preprocess(resnet50_x)\n",
    "\n",
    "mobilenetv2_x = tf.keras.utils.img_to_array(img)\n",
    "mobilenetv2_x = np.expand_dims(mobilenetv2_x, axis=0)\n",
    "mobilenetv2_x = mobilenetv2_preprocess(mobilenetv2_x)\n",
    "\n",
    "# Make predictions\n",
    "vgg16_preds = vgg16_model.predict(vgg16_x)\n",
    "resnet50_preds = resnet50_model.predict(resnet50_x)\n",
    "mobilenetv2_preds = mobilenetv2_model.predict(mobilenetv2_x)\n",
    "\n",
    "# Decode predictions\n",
    "vgg16_decoded = decode_predictions(vgg16_preds, top=3)[0]\n",
    "resnet50_decoded = decode_predictions(resnet50_preds, top=3)[0]\n",
    "mobilenetv2_decoded = decode_predictions(mobilenetv2_preds, top=3)[0]\n",
    "\n",
    "# Print top 3 results for each model\n",
    "print(\"VGG16:\")\n",
    "for label, class_id, confidence in vgg16_decoded:\n",
    "    print(f\"{label} ({class_id}): {confidence:.2f}\")\n",
    "\n",
    "print(\"\\nResNet50:\")\n",
    "for label, class_id, confidence in resnet50_decoded:\n",
    "    print(f\"{label} ({class_id}): {confidence:.2f}\")\n",
    "\n",
    "print(\"\\nMobileNetV2:\")\n",
    "for label, class_id, confidence in mobilenetv2_decoded:\n",
    "    print(f\"{label} ({class_id}): {confidence:.2f}\")\n",
    "\n",
    "# Compare top 3 results\n",
    "print(\"\\nComparison of top 3 results:\")\n",
    "for i in range(3):\n",
    "    vgg16_label = vgg16_decoded[i][0]\n",
    "    vgg16_class_id = vgg16_decoded[i][1]\n",
    "    resnet50_label = resnet50_decoded[i][0]\n",
    "    resnet50_class_id = resnet50_decoded[i][1]\n",
    "    mobilenetv2_label = mobilenetv2_decoded[i][0]\n",
    "    mobilenetv2_class_id = mobilenetv2_decoded[i][1]\n",
    "    print(f\"Rank {i+1}: VGG16={vgg16_label} ({vgg16_class_id}), ResNet50={resnet50_label} ({resnet50_class_id}), MobileNetV2={mobilenetv2_label} ({mobilenetv2_class_id})\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}