{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "s51pv6fmBigC"
},
"source": [
"Install mediapipe
"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "6QlneBN-Bna6"
},
"source": [
"Initiate path for silat Train Dataset + saving our mediapipe pose keypoints for later
\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"id": "08GvmxvUBtAM"
},
"outputs": [],
"source": [
"silat_train_images_dir = 'DATASET'\n",
"keypoint_outputs_dir = './coba/'"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "3qvWM7yjB28m"
},
"source": [
"Generate pose keypoints for each image in train dataset...
"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"id": "XaaPNUSzB69B"
},
"outputs": [],
"source": [
"import cv2\n",
"import numpy as np\n",
"import os\n",
"import tqdm\n",
"from mediapipe.python.solutions import drawing_utils as mp_drawing\n",
"from mediapipe.python.solutions import pose as mp_pose"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"pose_class_names = sorted([n for n in os.listdir(silat_train_images_dir)])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "gpgDUrEXB_cr"
},
"outputs": [],
"source": [
"for pose_class_name in pose_class_names:\n",
" image_names = sorted([n for n in os.listdir(os.path.join(silat_train_images_dir, pose_class_name))])\n",
"\n",
" try:\n",
" os.makedirs(os.path.join(keypoint_outputs_dir, pose_class_name))\n",
" except:\n",
" break\n",
"\n",
" print(\"Bootstrapping\", pose_class_name)\n",
" for image_name in tqdm.tqdm(image_names):\n",
" input_frame = cv2.imread(os.path.join(silat_train_images_dir, pose_class_name, image_name))\n",
" input_frame = cv2.cvtColor(input_frame, cv2.COLOR_BGR2RGB)\n",
"\n",
" with mp_pose.Pose() as pose_tracker:\n",
" result = pose_tracker.process(image=input_frame)\n",
" pose_landmarks = result.pose_landmarks\n",
"\n",
" output_frame = input_frame.copy()\n",
" mp_drawing.draw_landmarks(image=output_frame, landmark_list=pose_landmarks, connections=mp_pose.POSE_CONNECTIONS)\n",
"\n",
" output_frame = cv2.cvtColor(output_frame, cv2.COLOR_RGB2BGR)\n",
" # cv2.imwrite(os.path.join(train_outputs_dir, image_name), output_frame)\n",
"\n",
" if pose_landmarks is not None:\n",
" pose_landmarks = [[landmark.x, landmark.y, landmark.z] for landmark in pose_landmarks.landmark]\n",
" frame_height, frame_width = output_frame.shape[:2]\n",
"\n",
" \n",
" pose_landmarks *= np.array([frame_height, frame_height, frame_width])#untuk mengali x, y, z\n",
"\n",
" pose_landmarks = np.around(pose_landmarks, 5).flatten().astype(np.float32).tolist()\n",
"\n",
" npy_savepath = os.path.join(keypoint_outputs_dir, pose_class_name, image_name[0:-4]) # remove any .jpg, .png, etc suffix\n",
" np.save(npy_savepath, pose_landmarks)\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Bootstrapping A1_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 0%| | 0/366 [00:00Generate our train/test datasets"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"id": "DECcOWW5COH2"
},
"outputs": [],
"source": [
"from keras.utils import to_categorical\n",
"from sklearn.model_selection import train_test_split\n",
"from glob import glob"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"{'A1_benar': 0, 'A1_salah': 1, 'A2_benar': 2, 'A2_salah': 3, 'A3_benar': 4, 'A3_salah': 5, 'A4_benar': 6, 'A4_salah': 7, 'A5_benar': 8, 'A5_salah': 9, 'A6_benar': 10, 'A6_salah': 11, 'A7_benar': 12, 'A7_salah': 13}\n"
]
}
],
"source": [
"label_map = {label:num for num,label in enumerate(pose_class_names)}\n",
"print(label_map)"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"id": "wLc86ocBCUZa"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A1_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 362/362 [00:00<00:00, 626.28it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A1_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 221/221 [00:00<00:00, 661.62it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A2_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 268/268 [00:00<00:00, 584.76it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A2_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 275/275 [00:00<00:00, 838.58it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A3_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 244/244 [00:00<00:00, 1126.57it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A3_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 271/271 [00:00<00:00, 1133.06it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A4_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 224/224 [00:00<00:00, 821.87it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A4_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 300/300 [00:00<00:00, 1179.21it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A5_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 168/168 [00:00<00:00, 1180.30it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A5_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 205/205 [00:00<00:00, 1222.27it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A6_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 324/324 [00:00<00:00, 1021.77it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A6_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 186/186 [00:00<00:00, 1211.45it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A7_benar\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 212/212 [00:00<00:00, 1265.47it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"searching through A7_salah\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 77/77 [00:00<00:00, 682.64it/s]\n"
]
}
],
"source": [
"sequences, labels = [], []\n",
"\n",
"for pose_class_name in pose_class_names:\n",
" keypoint_names = glob(os.path.join(keypoint_outputs_dir, pose_class_name, \"*.npy\"))\n",
"\n",
" print(\"searching through {}\".format(pose_class_name))\n",
" for keypoint_name in tqdm.tqdm(keypoint_names):\n",
" file = np.load(keypoint_name)\n",
" sequences.append(file)\n",
" labels.append(label_map[pose_class_name])\n"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"id": "zR9J4mGmCXKu"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"3337\n",
"3337\n",
"(3337, 99)\n",
"(3337,)\n"
]
}
],
"source": [
"print(len(sequences))\n",
"print(len(labels))\n",
"\n",
"print(np.array(sequences).shape)\n",
"print(np.array(labels).shape)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"id": "cCUzP7sICgQD"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"(3337, 14)\n"
]
}
],
"source": [
"X = np.array(sequences)\n",
"Y = to_categorical(labels).astype(int)\n",
"print(Y.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "uMVxj9ewCmvB"
},
"source": [
"Preprocess Data"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"id": "7LG0Gfh9CqqY"
},
"outputs": [],
"source": [
"X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"id": "KTWZLczSCucm"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"(2669, 99)\n",
"668\n"
]
}
],
"source": [
"print(X_train.shape)\n",
"print(len(X_test))"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "2-HyyBmFCy84"
},
"source": [
"Generate our relatively simple but effective model...
"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"id": "XqCmWYcVC5Er"
},
"outputs": [],
"source": [
"from keras.models import Sequential\n",
"from keras.layers import LSTM, Dense, InputLayer, Dropout"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"id": "e43CovKMC8S4"
},
"outputs": [],
"source": [
"model = Sequential([\n",
" InputLayer(input_shape=(99,)),\n",
" Dense(32, activation='relu'),\n",
" Dense(64, activation='relu'),\n",
" Dense(128, activation='relu'),\n",
" Dense(14, activation='softmax')\n",
"])"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"id": "8KCKovQvC-jr"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Model: \"sequential\"\n",
"_________________________________________________________________\n",
" Layer (type) Output Shape Param # \n",
"=================================================================\n",
" dense (Dense) (None, 32) 3200 \n",
" \n",
" dense_1 (Dense) (None, 64) 2112 \n",
" \n",
" dense_2 (Dense) (None, 128) 8320 \n",
" \n",
" dense_3 (Dense) (None, 14) 1806 \n",
" \n",
"=================================================================\n",
"Total params: 15438 (60.30 KB)\n",
"Trainable params: 15438 (60.30 KB)\n",
"Non-trainable params: 0 (0.00 Byte)\n",
"_________________________________________________________________\n"
]
}
],
"source": [
"model.compile(optimizer=\"Adam\", loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
"model.summary()"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "DepzJ9dTDBdU"
},
"source": [
"Train our network!
"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"# Callback\n",
"\n",
"from tensorflow.keras.callbacks import Callback\n",
"class CustomCallback(Callback):\n",
" def on_epoch_end(self, epoch, logs=None):\n",
" if logs[\"categorical_accuracy\"] > 0.95 and logs[\"val_categorical_accuracy\"] > 0.95:\n",
" self.model.stop_training = True"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/150\n",
"84/84 [==============================] - 2s 6ms/step - loss: 59.5614 - categorical_accuracy: 0.2885 - val_loss: 18.8338 - val_categorical_accuracy: 0.4042\n",
"Epoch 2/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 10.3092 - categorical_accuracy: 0.5553 - val_loss: 6.8414 - val_categorical_accuracy: 0.5793\n",
"Epoch 3/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 6.3738 - categorical_accuracy: 0.6126 - val_loss: 5.2321 - val_categorical_accuracy: 0.6572\n",
"Epoch 4/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 4.7702 - categorical_accuracy: 0.6755 - val_loss: 4.7014 - val_categorical_accuracy: 0.6826\n",
"Epoch 5/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 4.0421 - categorical_accuracy: 0.6883 - val_loss: 3.5074 - val_categorical_accuracy: 0.7335\n",
"Epoch 6/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 3.4147 - categorical_accuracy: 0.7201 - val_loss: 3.9534 - val_categorical_accuracy: 0.6976\n",
"Epoch 7/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 2.7968 - categorical_accuracy: 0.7505 - val_loss: 3.5357 - val_categorical_accuracy: 0.6946\n",
"Epoch 8/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 2.5660 - categorical_accuracy: 0.7493 - val_loss: 4.6128 - val_categorical_accuracy: 0.6737\n",
"Epoch 9/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 2.4056 - categorical_accuracy: 0.7508 - val_loss: 2.6901 - val_categorical_accuracy: 0.7231\n",
"Epoch 10/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 2.0642 - categorical_accuracy: 0.7726 - val_loss: 2.4193 - val_categorical_accuracy: 0.7769\n",
"Epoch 11/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.6582 - categorical_accuracy: 0.7958 - val_loss: 3.5712 - val_categorical_accuracy: 0.6856\n",
"Epoch 12/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.6479 - categorical_accuracy: 0.7999 - val_loss: 2.1894 - val_categorical_accuracy: 0.7904\n",
"Epoch 13/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.4815 - categorical_accuracy: 0.8022 - val_loss: 2.4446 - val_categorical_accuracy: 0.7410\n",
"Epoch 14/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.2811 - categorical_accuracy: 0.8153 - val_loss: 2.1361 - val_categorical_accuracy: 0.7949\n",
"Epoch 15/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.4593 - categorical_accuracy: 0.8067 - val_loss: 2.1103 - val_categorical_accuracy: 0.7740\n",
"Epoch 16/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.3450 - categorical_accuracy: 0.8033 - val_loss: 2.1983 - val_categorical_accuracy: 0.7425\n",
"Epoch 17/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.1504 - categorical_accuracy: 0.8160 - val_loss: 2.5188 - val_categorical_accuracy: 0.7320\n",
"Epoch 18/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 1.2179 - categorical_accuracy: 0.8187 - val_loss: 1.6468 - val_categorical_accuracy: 0.7814\n",
"Epoch 19/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.0362 - categorical_accuracy: 0.8344 - val_loss: 1.6238 - val_categorical_accuracy: 0.7904\n",
"Epoch 20/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.9171 - categorical_accuracy: 0.8378 - val_loss: 2.1171 - val_categorical_accuracy: 0.7695\n",
"Epoch 21/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.9908 - categorical_accuracy: 0.8321 - val_loss: 1.5068 - val_categorical_accuracy: 0.8249\n",
"Epoch 22/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.2341 - categorical_accuracy: 0.8074 - val_loss: 1.4508 - val_categorical_accuracy: 0.8219\n",
"Epoch 23/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7555 - categorical_accuracy: 0.8531 - val_loss: 2.1852 - val_categorical_accuracy: 0.7605\n",
"Epoch 24/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.8684 - categorical_accuracy: 0.8460 - val_loss: 1.8595 - val_categorical_accuracy: 0.7799\n",
"Epoch 25/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.1621 - categorical_accuracy: 0.8145 - val_loss: 1.5206 - val_categorical_accuracy: 0.8278\n",
"Epoch 26/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.0637 - categorical_accuracy: 0.8250 - val_loss: 1.6952 - val_categorical_accuracy: 0.7949\n",
"Epoch 27/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.9199 - categorical_accuracy: 0.8291 - val_loss: 1.4013 - val_categorical_accuracy: 0.8189\n",
"Epoch 28/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.9212 - categorical_accuracy: 0.8486 - val_loss: 1.3321 - val_categorical_accuracy: 0.8383\n",
"Epoch 29/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7828 - categorical_accuracy: 0.8520 - val_loss: 1.1886 - val_categorical_accuracy: 0.8293\n",
"Epoch 30/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7386 - categorical_accuracy: 0.8468 - val_loss: 1.6173 - val_categorical_accuracy: 0.8114\n",
"Epoch 31/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7802 - categorical_accuracy: 0.8535 - val_loss: 1.5148 - val_categorical_accuracy: 0.7949\n",
"Epoch 32/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 1.0995 - categorical_accuracy: 0.8209 - val_loss: 1.6258 - val_categorical_accuracy: 0.8249\n",
"Epoch 33/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.8135 - categorical_accuracy: 0.8468 - val_loss: 1.5339 - val_categorical_accuracy: 0.7740\n",
"Epoch 34/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.8552 - categorical_accuracy: 0.8363 - val_loss: 1.3074 - val_categorical_accuracy: 0.8114\n",
"Epoch 35/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7019 - categorical_accuracy: 0.8662 - val_loss: 0.9738 - val_categorical_accuracy: 0.8518\n",
"Epoch 36/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7912 - categorical_accuracy: 0.8558 - val_loss: 1.4973 - val_categorical_accuracy: 0.7859\n",
"Epoch 37/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.7910 - categorical_accuracy: 0.8569 - val_loss: 1.4587 - val_categorical_accuracy: 0.8009\n",
"Epoch 38/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6530 - categorical_accuracy: 0.8629 - val_loss: 1.8864 - val_categorical_accuracy: 0.7769\n",
"Epoch 39/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.8420 - categorical_accuracy: 0.8531 - val_loss: 1.7212 - val_categorical_accuracy: 0.8114\n",
"Epoch 40/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7120 - categorical_accuracy: 0.8647 - val_loss: 1.4182 - val_categorical_accuracy: 0.8009\n",
"Epoch 41/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6591 - categorical_accuracy: 0.8670 - val_loss: 1.5737 - val_categorical_accuracy: 0.7844\n",
"Epoch 42/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6214 - categorical_accuracy: 0.8752 - val_loss: 1.3700 - val_categorical_accuracy: 0.8263\n",
"Epoch 43/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6163 - categorical_accuracy: 0.8707 - val_loss: 1.1052 - val_categorical_accuracy: 0.8189\n",
"Epoch 44/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6307 - categorical_accuracy: 0.8677 - val_loss: 0.9580 - val_categorical_accuracy: 0.8548\n",
"Epoch 45/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6735 - categorical_accuracy: 0.8651 - val_loss: 1.2040 - val_categorical_accuracy: 0.7934\n",
"Epoch 46/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.8484 - categorical_accuracy: 0.8516 - val_loss: 1.5534 - val_categorical_accuracy: 0.7455\n",
"Epoch 47/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.6835 - categorical_accuracy: 0.8629 - val_loss: 1.0494 - val_categorical_accuracy: 0.8443\n",
"Epoch 48/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6673 - categorical_accuracy: 0.8632 - val_loss: 1.0129 - val_categorical_accuracy: 0.8308\n",
"Epoch 49/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.5415 - categorical_accuracy: 0.8794 - val_loss: 0.9473 - val_categorical_accuracy: 0.8548\n",
"Epoch 50/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3883 - categorical_accuracy: 0.9033 - val_loss: 1.4681 - val_categorical_accuracy: 0.8114\n",
"Epoch 51/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7079 - categorical_accuracy: 0.8587 - val_loss: 1.2991 - val_categorical_accuracy: 0.8099\n",
"Epoch 52/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.5130 - categorical_accuracy: 0.8805 - val_loss: 1.0796 - val_categorical_accuracy: 0.8698\n",
"Epoch 53/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4199 - categorical_accuracy: 0.8973 - val_loss: 1.0935 - val_categorical_accuracy: 0.8234\n",
"Epoch 54/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.5201 - categorical_accuracy: 0.8846 - val_loss: 1.1886 - val_categorical_accuracy: 0.8353\n",
"Epoch 55/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.3947 - categorical_accuracy: 0.8985 - val_loss: 0.9999 - val_categorical_accuracy: 0.8443\n",
"Epoch 56/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.5866 - categorical_accuracy: 0.8779 - val_loss: 1.3747 - val_categorical_accuracy: 0.8054\n",
"Epoch 57/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6345 - categorical_accuracy: 0.8659 - val_loss: 1.4693 - val_categorical_accuracy: 0.8054\n",
"Epoch 58/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.9701 - categorical_accuracy: 0.8205 - val_loss: 2.3942 - val_categorical_accuracy: 0.7156\n",
"Epoch 59/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.7556 - categorical_accuracy: 0.8430 - val_loss: 1.0243 - val_categorical_accuracy: 0.8428\n",
"Epoch 60/150\n",
"84/84 [==============================] - 0s 6ms/step - loss: 0.4747 - categorical_accuracy: 0.8805 - val_loss: 1.2611 - val_categorical_accuracy: 0.8204\n",
"Epoch 61/150\n",
"84/84 [==============================] - 1s 10ms/step - loss: 0.5938 - categorical_accuracy: 0.8764 - val_loss: 0.9473 - val_categorical_accuracy: 0.8653\n",
"Epoch 62/150\n",
"84/84 [==============================] - 1s 9ms/step - loss: 0.4280 - categorical_accuracy: 0.8973 - val_loss: 0.8654 - val_categorical_accuracy: 0.8593\n",
"Epoch 63/150\n",
"84/84 [==============================] - 1s 6ms/step - loss: 0.5382 - categorical_accuracy: 0.8831 - val_loss: 1.2460 - val_categorical_accuracy: 0.7964\n",
"Epoch 64/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4442 - categorical_accuracy: 0.8839 - val_loss: 1.1299 - val_categorical_accuracy: 0.8338\n",
"Epoch 65/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6336 - categorical_accuracy: 0.8640 - val_loss: 1.2046 - val_categorical_accuracy: 0.8159\n",
"Epoch 66/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4335 - categorical_accuracy: 0.8883 - val_loss: 1.1096 - val_categorical_accuracy: 0.8338\n",
"Epoch 67/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4869 - categorical_accuracy: 0.8809 - val_loss: 0.7997 - val_categorical_accuracy: 0.8473\n",
"Epoch 68/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3781 - categorical_accuracy: 0.8988 - val_loss: 0.9846 - val_categorical_accuracy: 0.8219\n",
"Epoch 69/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4115 - categorical_accuracy: 0.8958 - val_loss: 0.8778 - val_categorical_accuracy: 0.8428\n",
"Epoch 70/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.6819 - categorical_accuracy: 0.8483 - val_loss: 1.1496 - val_categorical_accuracy: 0.8353\n",
"Epoch 71/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.4487 - categorical_accuracy: 0.8921 - val_loss: 0.9183 - val_categorical_accuracy: 0.8623\n",
"Epoch 72/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3784 - categorical_accuracy: 0.9052 - val_loss: 0.8161 - val_categorical_accuracy: 0.8293\n",
"Epoch 73/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3625 - categorical_accuracy: 0.9015 - val_loss: 0.9973 - val_categorical_accuracy: 0.8278\n",
"Epoch 74/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.5614 - categorical_accuracy: 0.8670 - val_loss: 1.1763 - val_categorical_accuracy: 0.7964\n",
"Epoch 75/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4715 - categorical_accuracy: 0.8797 - val_loss: 0.9896 - val_categorical_accuracy: 0.8503\n",
"Epoch 76/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4352 - categorical_accuracy: 0.8887 - val_loss: 0.9062 - val_categorical_accuracy: 0.8623\n",
"Epoch 77/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3908 - categorical_accuracy: 0.9011 - val_loss: 0.8319 - val_categorical_accuracy: 0.8578\n",
"Epoch 78/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4396 - categorical_accuracy: 0.8861 - val_loss: 0.7948 - val_categorical_accuracy: 0.8608\n",
"Epoch 79/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.3585 - categorical_accuracy: 0.9060 - val_loss: 1.2951 - val_categorical_accuracy: 0.8189\n",
"Epoch 80/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4548 - categorical_accuracy: 0.8809 - val_loss: 1.1766 - val_categorical_accuracy: 0.8144\n",
"Epoch 81/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4061 - categorical_accuracy: 0.8951 - val_loss: 0.7043 - val_categorical_accuracy: 0.8533\n",
"Epoch 82/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.4465 - categorical_accuracy: 0.8801 - val_loss: 0.8877 - val_categorical_accuracy: 0.8338\n",
"Epoch 83/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3088 - categorical_accuracy: 0.9120 - val_loss: 0.8085 - val_categorical_accuracy: 0.8578\n",
"Epoch 84/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2667 - categorical_accuracy: 0.9153 - val_loss: 0.8084 - val_categorical_accuracy: 0.8623\n",
"Epoch 85/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3052 - categorical_accuracy: 0.9161 - val_loss: 0.8641 - val_categorical_accuracy: 0.8383\n",
"Epoch 86/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2929 - categorical_accuracy: 0.9209 - val_loss: 0.6837 - val_categorical_accuracy: 0.8668\n",
"Epoch 87/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.3432 - categorical_accuracy: 0.9063 - val_loss: 0.8765 - val_categorical_accuracy: 0.8563\n",
"Epoch 88/150\n",
"84/84 [==============================] - 0s 6ms/step - loss: 0.3926 - categorical_accuracy: 0.8947 - val_loss: 0.8860 - val_categorical_accuracy: 0.8308\n",
"Epoch 89/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.6399 - categorical_accuracy: 0.8449 - val_loss: 0.9501 - val_categorical_accuracy: 0.8293\n",
"Epoch 90/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.4665 - categorical_accuracy: 0.8760 - val_loss: 1.0812 - val_categorical_accuracy: 0.8069\n",
"Epoch 91/150\n",
"84/84 [==============================] - 0s 2ms/step - loss: 0.6251 - categorical_accuracy: 0.8580 - val_loss: 0.8070 - val_categorical_accuracy: 0.8548\n",
"Epoch 92/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.3402 - categorical_accuracy: 0.9011 - val_loss: 0.6712 - val_categorical_accuracy: 0.8668\n",
"Epoch 93/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.3920 - categorical_accuracy: 0.8936 - val_loss: 0.7727 - val_categorical_accuracy: 0.8593\n",
"Epoch 94/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.3755 - categorical_accuracy: 0.8902 - val_loss: 0.7758 - val_categorical_accuracy: 0.8847\n",
"Epoch 95/150\n",
"84/84 [==============================] - 1s 6ms/step - loss: 0.3214 - categorical_accuracy: 0.9052 - val_loss: 0.7342 - val_categorical_accuracy: 0.8563\n",
"Epoch 96/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2908 - categorical_accuracy: 0.9157 - val_loss: 0.7648 - val_categorical_accuracy: 0.8608\n",
"Epoch 97/150\n",
"84/84 [==============================] - 1s 6ms/step - loss: 0.3505 - categorical_accuracy: 0.8992 - val_loss: 0.7152 - val_categorical_accuracy: 0.8698\n",
"Epoch 98/150\n",
"84/84 [==============================] - 1s 9ms/step - loss: 0.2748 - categorical_accuracy: 0.9209 - val_loss: 0.7209 - val_categorical_accuracy: 0.8578\n",
"Epoch 99/150\n",
"84/84 [==============================] - 0s 6ms/step - loss: 0.2692 - categorical_accuracy: 0.9228 - val_loss: 0.8454 - val_categorical_accuracy: 0.8698\n",
"Epoch 100/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2843 - categorical_accuracy: 0.9172 - val_loss: 0.7282 - val_categorical_accuracy: 0.8623\n",
"Epoch 101/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.3704 - categorical_accuracy: 0.9082 - val_loss: 0.6408 - val_categorical_accuracy: 0.8638\n",
"Epoch 102/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2191 - categorical_accuracy: 0.9269 - val_loss: 0.6216 - val_categorical_accuracy: 0.8877\n",
"Epoch 103/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2495 - categorical_accuracy: 0.9269 - val_loss: 0.7032 - val_categorical_accuracy: 0.8488\n",
"Epoch 104/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.4176 - categorical_accuracy: 0.8891 - val_loss: 0.7150 - val_categorical_accuracy: 0.8743\n",
"Epoch 105/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.1783 - categorical_accuracy: 0.9408 - val_loss: 0.6601 - val_categorical_accuracy: 0.8668\n",
"Epoch 106/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.2179 - categorical_accuracy: 0.9322 - val_loss: 0.6447 - val_categorical_accuracy: 0.8728\n",
"Epoch 107/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.2955 - categorical_accuracy: 0.9093 - val_loss: 0.8380 - val_categorical_accuracy: 0.8174\n",
"Epoch 108/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.3313 - categorical_accuracy: 0.9045 - val_loss: 0.7468 - val_categorical_accuracy: 0.8548\n",
"Epoch 109/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.2961 - categorical_accuracy: 0.9045 - val_loss: 0.6795 - val_categorical_accuracy: 0.8443\n",
"Epoch 110/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.4318 - categorical_accuracy: 0.8857 - val_loss: 1.3955 - val_categorical_accuracy: 0.7769\n",
"Epoch 111/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.7229 - categorical_accuracy: 0.8321 - val_loss: 0.8247 - val_categorical_accuracy: 0.8204\n",
"Epoch 112/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.3931 - categorical_accuracy: 0.8850 - val_loss: 0.6159 - val_categorical_accuracy: 0.8593\n",
"Epoch 113/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.2134 - categorical_accuracy: 0.9303 - val_loss: 0.7091 - val_categorical_accuracy: 0.8743\n",
"Epoch 114/150\n",
"84/84 [==============================] - 0s 3ms/step - loss: 0.2762 - categorical_accuracy: 0.9284 - val_loss: 0.6848 - val_categorical_accuracy: 0.8668\n",
"Epoch 115/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2964 - categorical_accuracy: 0.9097 - val_loss: 0.8741 - val_categorical_accuracy: 0.8428\n",
"Epoch 116/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.4242 - categorical_accuracy: 0.8764 - val_loss: 0.6144 - val_categorical_accuracy: 0.8772\n",
"Epoch 117/150\n",
"84/84 [==============================] - 1s 9ms/step - loss: 0.2758 - categorical_accuracy: 0.9161 - val_loss: 0.9470 - val_categorical_accuracy: 0.8383\n",
"Epoch 118/150\n",
"84/84 [==============================] - 1s 11ms/step - loss: 0.3158 - categorical_accuracy: 0.9007 - val_loss: 0.5806 - val_categorical_accuracy: 0.8698\n",
"Epoch 119/150\n",
"84/84 [==============================] - 0s 6ms/step - loss: 0.3320 - categorical_accuracy: 0.9101 - val_loss: 0.7224 - val_categorical_accuracy: 0.8518\n",
"Epoch 120/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2426 - categorical_accuracy: 0.9176 - val_loss: 0.5684 - val_categorical_accuracy: 0.8922\n",
"Epoch 121/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2339 - categorical_accuracy: 0.9296 - val_loss: 0.5851 - val_categorical_accuracy: 0.8832\n",
"Epoch 122/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1627 - categorical_accuracy: 0.9442 - val_loss: 0.6781 - val_categorical_accuracy: 0.8473\n",
"Epoch 123/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2406 - categorical_accuracy: 0.9262 - val_loss: 0.5436 - val_categorical_accuracy: 0.8787\n",
"Epoch 124/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.4101 - categorical_accuracy: 0.8816 - val_loss: 0.7285 - val_categorical_accuracy: 0.8413\n",
"Epoch 125/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.3947 - categorical_accuracy: 0.8835 - val_loss: 0.6861 - val_categorical_accuracy: 0.8368\n",
"Epoch 126/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2936 - categorical_accuracy: 0.9131 - val_loss: 0.5750 - val_categorical_accuracy: 0.8757\n",
"Epoch 127/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2258 - categorical_accuracy: 0.9299 - val_loss: 0.7586 - val_categorical_accuracy: 0.8473\n",
"Epoch 128/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.3457 - categorical_accuracy: 0.8970 - val_loss: 0.6622 - val_categorical_accuracy: 0.8668\n",
"Epoch 129/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2254 - categorical_accuracy: 0.9221 - val_loss: 0.6610 - val_categorical_accuracy: 0.8713\n",
"Epoch 130/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3175 - categorical_accuracy: 0.9063 - val_loss: 0.5772 - val_categorical_accuracy: 0.8608\n",
"Epoch 131/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1580 - categorical_accuracy: 0.9423 - val_loss: 0.5691 - val_categorical_accuracy: 0.8802\n",
"Epoch 132/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.1503 - categorical_accuracy: 0.9479 - val_loss: 0.6099 - val_categorical_accuracy: 0.8802\n",
"Epoch 133/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1663 - categorical_accuracy: 0.9412 - val_loss: 0.5149 - val_categorical_accuracy: 0.8907\n",
"Epoch 134/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1419 - categorical_accuracy: 0.9543 - val_loss: 0.4774 - val_categorical_accuracy: 0.9057\n",
"Epoch 135/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1535 - categorical_accuracy: 0.9460 - val_loss: 0.6644 - val_categorical_accuracy: 0.8593\n",
"Epoch 136/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1955 - categorical_accuracy: 0.9341 - val_loss: 0.5563 - val_categorical_accuracy: 0.8847\n",
"Epoch 137/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2206 - categorical_accuracy: 0.9318 - val_loss: 0.4799 - val_categorical_accuracy: 0.9027\n",
"Epoch 138/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.1952 - categorical_accuracy: 0.9314 - val_loss: 0.6195 - val_categorical_accuracy: 0.8967\n",
"Epoch 139/150\n",
"84/84 [==============================] - 1s 9ms/step - loss: 0.2455 - categorical_accuracy: 0.9217 - val_loss: 0.7124 - val_categorical_accuracy: 0.8398\n",
"Epoch 140/150\n",
"84/84 [==============================] - 1s 7ms/step - loss: 0.2214 - categorical_accuracy: 0.9266 - val_loss: 0.6361 - val_categorical_accuracy: 0.8668\n",
"Epoch 141/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2680 - categorical_accuracy: 0.9191 - val_loss: 0.6368 - val_categorical_accuracy: 0.8623\n",
"Epoch 142/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.3060 - categorical_accuracy: 0.9071 - val_loss: 0.5976 - val_categorical_accuracy: 0.8713\n",
"Epoch 143/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.1961 - categorical_accuracy: 0.9333 - val_loss: 0.5137 - val_categorical_accuracy: 0.8847\n",
"Epoch 144/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2862 - categorical_accuracy: 0.9116 - val_loss: 0.7269 - val_categorical_accuracy: 0.8593\n",
"Epoch 145/150\n",
"84/84 [==============================] - 0s 6ms/step - loss: 0.1872 - categorical_accuracy: 0.9397 - val_loss: 0.4624 - val_categorical_accuracy: 0.8982\n",
"Epoch 146/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.1568 - categorical_accuracy: 0.9513 - val_loss: 0.7563 - val_categorical_accuracy: 0.8473\n",
"Epoch 147/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2408 - categorical_accuracy: 0.9232 - val_loss: 0.6378 - val_categorical_accuracy: 0.8743\n",
"Epoch 148/150\n",
"84/84 [==============================] - 0s 5ms/step - loss: 0.2037 - categorical_accuracy: 0.9314 - val_loss: 1.2473 - val_categorical_accuracy: 0.7964\n",
"Epoch 149/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.4259 - categorical_accuracy: 0.8913 - val_loss: 0.5930 - val_categorical_accuracy: 0.8862\n",
"Epoch 150/150\n",
"84/84 [==============================] - 0s 4ms/step - loss: 0.2070 - categorical_accuracy: 0.9326 - val_loss: 0.7578 - val_categorical_accuracy: 0.8503\n"
]
}
],
"source": [
"history = model.fit(X_train, Y_train,\n",
" validation_data=(X_test, Y_test),\n",
" epochs=150,\n",
" callbacks=[CustomCallback()])"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"id": "IA2Qz74HDEV2"
},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"\n",
"plt.figure(figsize=(12, 4))\n",
"\n",
"# Plot Akurasi\n",
"plt.subplot(1, 2, 1)\n",
"plt.plot(history.history['categorical_accuracy'], label='Train Accuracy')\n",
"plt.plot(history.history['val_categorical_accuracy'], label='Validation Accuracy')\n",
"plt.xlabel('Epoch')\n",
"plt.ylabel('Accuracy')\n",
"plt.legend()\n",
"plt.title('Training vs Validation Accuracy')\n",
"\n",
"# Plot Loss\n",
"plt.subplot(1, 2, 2)\n",
"plt.plot(history.history['loss'], label='Train Loss')\n",
"plt.plot(history.history['val_loss'], label='Validation Loss')\n",
"plt.xlabel('Epoch')\n",
"plt.ylabel('Loss')\n",
"plt.legend()\n",
"plt.title('Training vs Validation Loss')\n",
"\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"21/21 [==============================] - 0s 3ms/step - loss: 0.7578 - categorical_accuracy: 0.8503\n",
"Loss pada data uji: 0.76\n",
"Akurasi pada data uji: 0.85\n"
]
}
],
"source": [
"# Evaluasi model dengan data uji\n",
"loss, accuracy = model.evaluate(X_test, Y_test)\n",
"\n",
"# Tampilkan hasil evaluasi\n",
"print(f\"Loss pada data uji: {loss:.2f}\")\n",
"print(f\"Akurasi pada data uji: {accuracy:.2f}\")\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "z3SZgsEiDHcM"
},
"source": [
"Do prediction
"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"poses_mapping = {'A1_benar': 0, 'A1_salah': 1,'A2_benar': 2, 'A2_salah':3, 'A3_benar': 4, 'A3_salah':5, 'A4_benar':6, 'A4_salah':7, 'A5_benar':8, 'A5_salah':9, 'A6_benar':10, 'A6_salah':11, 'A7_benar':12, 'A7_salah':13 }"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"21/21 [==============================] - 0s 2ms/step\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from sklearn.metrics import confusion_matrix\n",
"import seaborn as sns\n",
"\n",
"# Prediksi kelas untuk data uji\n",
"Y_pred = model.predict(X_test)\n",
"y_pred_classes = np.argmax(Y_pred, axis=1)\n",
"y_true = np.argmax(Y_test, axis=1)\n",
"\n",
"# Mendapatkan matriks kebingungan\n",
"confusion_mtx = confusion_matrix(y_true, y_pred_classes)\n",
"\n",
"# Menampilkan matriks kebingungan dengan heatmap\n",
"plt.figure(figsize=(10, 8))\n",
"sns.heatmap(confusion_mtx, annot=True, fmt='d', cmap='Blues', \n",
" xticklabels=poses_mapping, yticklabels=poses_mapping)\n",
"plt.xlabel('Predicted')\n",
"plt.ylabel('True')\n",
"plt.title('Confusion Matrix')\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [],
"source": [
"poses_label = {'A1_benar': 0, 'A1_salah': 1,'A2_benar': 2, 'A2_salah':3, 'A3_benar': 4, 'A3_salah':5, 'A4_benar':6, 'A4_salah':7, 'A5_benar':8, 'A5_salah':9, 'A6_benar':10, 'A6_salah':11, 'A7_benar':12, 'A7_salah':13 }"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" precision recall f1-score support\n",
"\n",
" A1_benar 0.76 0.97 0.86 79\n",
" A1_salah 0.82 0.85 0.84 39\n",
" A2_benar 0.69 0.83 0.75 48\n",
" A2_salah 0.80 0.94 0.86 64\n",
" A3_benar 0.89 0.65 0.75 51\n",
" A3_salah 0.89 0.92 0.90 36\n",
" A4_benar 0.82 0.66 0.73 41\n",
" A4_salah 0.91 0.87 0.89 60\n",
" A5_benar 0.88 0.74 0.80 38\n",
" A5_salah 1.00 0.83 0.91 59\n",
" A6_benar 0.91 0.84 0.88 58\n",
" A6_salah 0.89 0.97 0.93 35\n",
" A7_benar 0.93 0.88 0.90 43\n",
" A7_salah 0.94 0.88 0.91 17\n",
"\n",
" accuracy 0.85 668\n",
" macro avg 0.87 0.85 0.85 668\n",
"weighted avg 0.86 0.85 0.85 668\n",
"\n"
]
}
],
"source": [
"from sklearn.metrics import classification_report\n",
"classification_rep = classification_report(y_true, y_pred_classes, target_names=poses_label)\n",
"print(classification_rep)"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {
"id": "DFaaZIERDM9Z"
},
"outputs": [],
"source": [
"# poses_mapping = {'A1': 0, 'A2': 1, 'A3': 2, 'A4': 3, 'A5': 4, 'A6': 5, 'A7': 6}\n",
"# result = model.predict(X_test)\n",
"\n",
"# print(\"Predicted pose:\", poses_mapping[np.argmax(result[0])])\n",
"# print(\"Actual pose:\", poses_mapping[np.argmax(Y_test[0])])"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "Bag6q0iCDP62"
},
"source": [
"Save the model
"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {
"id": "9XDvY8SqDUbJ"
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"c:\\Users\\Windows 10\\AppData\\Local\\Programs\\Python\\Python311\\Lib\\site-packages\\keras\\src\\engine\\training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
" saving_api.save_model(\n"
]
}
],
"source": [
"model.save('./model_save/modelnn.h5')"
]
}
],
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"display_name": "Python 3",
"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": "3.11.4"
}
},
"nbformat": 4,
"nbformat_minor": 0
}