Merge branch 'main' into pages

src/HowBadIsMyBatch.ipynb

@@ -220,6 +220,105 @@
     "        time.sleep(2)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d9b72506",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import tensorflow as tf\n",
+    "from tensorflow import keras\n",
+    "from tensorflow.keras import layers\n",
+    "from PIL import Image\n",
+    "import numpy as np\n",
+    "import io\n",
+    "\n",
+    "# copied from value of characters variable in captcha_ocr.ipynb or captcha_ocr_trainAndSaveModel.ipynb\n",
+    "characters = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F', 'a', 'b', 'c', 'd', 'e', 'f']\n",
+    "\n",
+    "batch_size = 1\n",
+    "\n",
+    "img_width = 241\n",
+    "img_height = 62\n",
+    "\n",
+    "downsample_factor = 4\n",
+    "\n",
+    "# copied from value of max_length variable in captcha_ocr.ipynb or captcha_ocr_trainAndSaveModel.ipynb\n",
+    "max_length = 6\n",
+    "\n",
+    "char_to_num = layers.StringLookup(\n",
+    "    vocabulary=list(characters),\n",
+    "    mask_token=None)\n",
+    "\n",
+    "num_to_char = layers.StringLookup(\n",
+    "    vocabulary=char_to_num.get_vocabulary(),\n",
+    "    mask_token=None, invert=True)\n",
+    "\n",
+    "\n",
+    "def encode_single_sample(img_path):\n",
+    "    # 1. Read image\n",
+    "    img = tf.io.read_file(img_path)\n",
+    "    # 2. Decode and convert to grayscale\n",
+    "    img = tf.io.decode_png(img, channels=1)\n",
+    "    # 3. Convert to float32 in [0, 1] range\n",
+    "    img = tf.image.convert_image_dtype(img, tf.float32)\n",
+    "    # 4. Resize to the desired size\n",
+    "    img = tf.image.resize(img, [img_height, img_width])\n",
+    "    # 5. Transpose the image because we want the time\n",
+    "    # dimension to correspond to the width of the image.\n",
+    "    img = tf.transpose(img, perm=[1, 0, 2])\n",
+    "    # 7. Return a dict as our model is expecting two inputs\n",
+    "    return img\n",
+    "\n",
+    "\n",
+    "def decode_batch_predictions(pred):\n",
+    "    input_len = np.ones(pred.shape[0]) * pred.shape[1]\n",
+    "    # Use greedy search. For complex tasks, you can use beam search\n",
+    "    results = keras.backend.ctc_decode(pred, input_length=input_len, greedy=True)[0][0][:, :max_length]\n",
+    "    # Iterate over the results and get back the text\n",
+    "    output_text = []\n",
+    "    for res in results:\n",
+    "        res = tf.strings.reduce_join(num_to_char(res)).numpy().decode(\"utf-8\")\n",
+    "        output_text.append(res)\n",
+    "    return output_text\n",
+    "\n",
+    "\n",
+    "def load_model():\n",
+    "    _model = keras.models.load_model('model')\n",
+    "    model = keras.models.Model(\n",
+    "        _model.get_layer(name=\"image\").input,\n",
+    "        _model.get_layer(name=\"dense2\").output)\n",
+    "    return model\n",
+    "\n",
+    "\n",
+    "def getTextInCaptchaImage(captchaImageFile):\n",
+    "    batchImages = getBatchImagesFromFile(captchaImageFile)\n",
+    "    preds = model.predict(batchImages)\n",
+    "    return decode_batch_predictions(preds)[0]\n",
+    "\n",
+    "\n",
+    "def getBatchImagesFromFile(imageFile):\n",
+    "    return list(asDataset(imageFile).as_numpy_iterator())[0]\n",
+    "\n",
+    "\n",
+    "def asDataset(imageFile):\n",
+    "    dataset = tf.data.Dataset.from_tensor_slices([imageFile])\n",
+    "    dataset = (\n",
+    "        dataset\n",
+    "        .map(encode_single_sample, num_parallel_calls=tf.data.AUTOTUNE)\n",
+    "        .batch(batch_size)\n",
+    "        .prefetch(buffer_size=tf.data.AUTOTUNE)\n",
+    "    )\n",
+    "    return dataset\n",
+    "\n",
+    "\n",
+    "print(\"loading model...\")\n",
+    "model = load_model()\n",
+    "model.summary()"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -229,16 +328,16 @@
    "source": [
     "import os\n",
     "\n",
-    "def getTextInCaptchaImage(captchaImageFile):\n",
-    "    baseDir = \"~/AndroidStudioProjects/TextRecognizer\"\n",
-    "    ! cp $captchaImageFile $baseDir/app/src/main/assets/captchas/captcha_image.jpeg\n",
-    "    ! cd $baseDir;./gradlew connectedAndroidTest\n",
-    "    textInCaptchaImage = ! adb shell \"run-as org.textrecognizer cat /data/data/org.textrecognizer/files/captcha_image.txt\"\n",
-    "    return textInCaptchaImage[0]\n",
+    "#def getTextInCaptchaImage(captchaImageFile):\n",
+    "#    baseDir = \"~/AndroidStudioProjects/TextRecognizer\"\n",
+    "#    ! cp $captchaImageFile $baseDir/app/src/main/assets/captchas/captcha_image.jpeg\n",
+    "#    ! cd $baseDir;./gradlew connectedAndroidTest\n",
+    "#    textInCaptchaImage = ! adb shell \"run-as org.textrecognizer cat /data/data/org.textrecognizer/files/captcha_image.txt\"\n",
+    "#    return textInCaptchaImage[0]\n",
     "    \n",
     "def solveCaptchaAndStartFileDownload(driver, captchaImageFile):\n",
     "    saveCaptchaImageAs(driver, captchaImageFile)\n",
-    "    textInCaptchaImage = AndroidEmulator.run(lambda: getTextInCaptchaImage(captchaImageFile))\n",
+    "    textInCaptchaImage = getTextInCaptchaImage(captchaImageFile)\n",
     "    display('textInCaptchaImage: ', textInCaptchaImage)\n",
     "    driver.find_element(By.ID, \"verificationCode\").send_keys(textInCaptchaImage)\n",
     "    driver.find_element(By.CSS_SELECTOR, '[name=\"downloadbut\"]').click()\n",