train/fitness_model_trainer.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "fitness_model_trainer.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oJOM6b2Qzg5y"
      },
      "source": [
        "# Training a model from fitness data\n",
        "\n",
        "This code was adapted from the tutorial 'Get started with machine learning on Arduino' by Sandeep Mistry & Dominic Pajak in the Arduino blog. Special thanks for the Arduino team for giving the resources as a starting point-\n",
        "[Link to the blog](https://blog.arduino.cc/2019/10/15/get-started-with-machine-learning-on-arduino/)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Il1VWxc1XMr"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MIZQ4GYGusaS",
        "outputId": "1ee626a0-2475-416f-a139-74842d8ddb67"
      },
      "source": [
        "print(\"numpy:\", np.__version__)\n",
        "print(\"pandas:\", pd.__version__)\n",
        "print(\"tensorflow:\", tf.__version__)\n",
        "print(\"seaborn\", sns.__version__)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "numpy: 1.19.5\n",
            "pandas: 1.1.5\n",
            "tensorflow: 2.5.0\n",
            "seaborn 0.11.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZNV2tl3omicB"
      },
      "source": [
        "# Define constants required for program\n",
        "np.set_printoptions(precision=3, suppress=True)\n",
        "\n",
        "SEED = 1337\n",
        "np.random.seed(SEED)\n",
        "tf.random.set_seed(SEED)\n",
        "\n",
        "ACTIONS = [\"walking\", \"sitting\", \"standing\"]\n",
        "SAMPLES_PER_ACTION = 50\n",
        "NUM_ACTIONS = len(ACTIONS)\n",
        "ONE_HOT_ENCODED_ACTIONS = np.eye(NUM_ACTIONS)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JbDRR90Nmw3C",
        "outputId": "92bfd7b8-2df8-4a60-999b-e3d6dc28f2d6"
      },
      "source": [
        "# Parse the data\n",
        "inputs = []\n",
        "outputs = []\n",
        "\n",
        "for i in range(NUM_ACTIONS):\n",
        "  action = ACTIONS[i]\n",
        "  print(f\"Processing file for action '{action}'.\")\n",
        "\n",
        "  output = ONE_HOT_ENCODED_ACTIONS[i]\n",
        "  df = pd.read_csv(action+\".csv\", skip_blank_lines=True)\n",
        "  num_recordings = int(df.shape[0] / SAMPLES_PER_ACTION)\n",
        "  print(f\"\\tThere are {num_recordings} recordings in the {action} file.\")\n",
        "\n",
        "  for i in range(num_recordings):\n",
        "    tensor = []\n",
        "    for j in range(SAMPLES_PER_ACTION):\n",
        "      index = i * SAMPLES_PER_ACTION + j\n",
        "      tensor += [\n",
        "          (df['acc.x'][index]) / 4,\n",
        "          (df['acc.y'][index]) / 4,\n",
        "          (df['acc.z'][index]) / 4,\n",
        "          (df['gyro.x'][index]) / 2000,\n",
        "          (df['gyro.y'][index]) / 2000,\n",
        "          (df['gyro.z'][index]) / 2000\n",
        "      ]\n",
        "\n",
        "    inputs.append(tensor)\n",
        "    outputs.append(output)\n",
        "\n",
        "inputs = np.array(inputs)\n",
        "outputs = np.array(outputs)\n",
        "\n",
        "print(\"Data set parsing and preparation complete.\")"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Processing file for action 'walking'.\n",
            "\tThere are 30 recordings in the walking file.\n",
            "Processing file for action 'sitting'.\n",
            "\tThere are 30 recordings in the sitting file.\n",
            "Processing file for action 'standing'.\n",
            "\tThere are 30 recordings in the standing file.\n",
            "Data set parsing and preparation complete.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "S_s6ysDlm1vG",
        "outputId": "05588e1f-eef2-4103-ddb2-ae4c472f7c3a"
      },
      "source": [
        "# Randomize the order of the data and split into train, test and validation sets\n",
        "num_inputs = len(inputs)\n",
        "randomize = np.arange(num_inputs)\n",
        "np.random.shuffle(randomize)\n",
        "\n",
        "inputs = inputs[randomize]\n",
        "outputs = outputs[randomize]\n",
        "\n",
        "TRAIN_SPLIT = int(0.6 * num_inputs)\n",
        "TEST_SPLIT = int(0.2 * num_inputs + TRAIN_SPLIT)\n",
        "\n",
        "inputs_train, inputs_test, inputs_validate = np.split(inputs, [TRAIN_SPLIT, TEST_SPLIT])\n",
        "outputs_train, outputs_test, outputs_validate = np.split(outputs, [TRAIN_SPLIT, TEST_SPLIT])\n",
        "\n",
        "print(\"Data set randomization and splitting complete.\")"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Data set randomization and splitting complete.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iTuRoXuTnKHd"
      },
      "source": [
        "# Build model\n",
        "model = tf.keras.Sequential()\n",
        "model.add(tf.keras.layers.Dense(50, activation='relu'))\n",
        "model.add(tf.keras.layers.Dense(15, activation='relu'))\n",
        "model.add(tf.keras.layers.Dense(NUM_ACTIONS, activation='softmax'))\n",
        "model.compile(optimizer='rmsprop', loss='mse', metrics=['accuracy'])\n"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kAM_q-XTnWxZ",
        "outputId": "dee4027f-54da-4954-fe72-ee42756fa3d3"
      },
      "source": [
        "# Train model\n",
        "history = model.fit(inputs_train, outputs_train, epochs=100, batch_size=1, validation_data=(inputs_validate, outputs_validate))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "54/54 [==============================] - 1s 5ms/step - loss: 0.2094 - accuracy: 0.6481 - val_loss: 0.2045 - val_accuracy: 0.7778\n",
            "Epoch 2/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.1746 - accuracy: 0.7963 - val_loss: 0.1710 - val_accuracy: 1.0000\n",
            "Epoch 3/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.1353 - accuracy: 0.9444 - val_loss: 0.1422 - val_accuracy: 1.0000\n",
            "Epoch 4/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.1034 - accuracy: 0.9630 - val_loss: 0.1086 - val_accuracy: 1.0000\n",
            "Epoch 5/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0784 - accuracy: 0.9815 - val_loss: 0.0811 - val_accuracy: 1.0000\n",
            "Epoch 6/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0604 - accuracy: 0.9630 - val_loss: 0.0581 - val_accuracy: 1.0000\n",
            "Epoch 7/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0465 - accuracy: 0.9815 - val_loss: 0.0376 - val_accuracy: 1.0000\n",
            "Epoch 8/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0383 - accuracy: 0.9630 - val_loss: 0.0261 - val_accuracy: 1.0000\n",
            "Epoch 9/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0311 - accuracy: 0.9444 - val_loss: 0.0191 - val_accuracy: 1.0000\n",
            "Epoch 10/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0272 - accuracy: 0.9815 - val_loss: 0.0135 - val_accuracy: 1.0000\n",
            "Epoch 11/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0243 - accuracy: 0.9444 - val_loss: 0.0130 - val_accuracy: 1.0000\n",
            "Epoch 12/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0237 - accuracy: 0.9630 - val_loss: 0.0081 - val_accuracy: 1.0000\n",
            "Epoch 13/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0220 - accuracy: 0.9630 - val_loss: 0.0068 - val_accuracy: 1.0000\n",
            "Epoch 14/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0211 - accuracy: 0.9630 - val_loss: 0.0057 - val_accuracy: 1.0000\n",
            "Epoch 15/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0183 - accuracy: 0.9815 - val_loss: 0.0047 - val_accuracy: 1.0000\n",
            "Epoch 16/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0179 - accuracy: 0.9815 - val_loss: 0.0038 - val_accuracy: 1.0000\n",
            "Epoch 17/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0168 - accuracy: 0.9815 - val_loss: 0.0027 - val_accuracy: 1.0000\n",
            "Epoch 18/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0181 - accuracy: 0.9815 - val_loss: 0.0029 - val_accuracy: 1.0000\n",
            "Epoch 19/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0164 - accuracy: 0.9815 - val_loss: 0.0022 - val_accuracy: 1.0000\n",
            "Epoch 20/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0162 - accuracy: 0.9815 - val_loss: 0.0021 - val_accuracy: 1.0000\n",
            "Epoch 21/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0150 - accuracy: 0.9815 - val_loss: 0.0015 - val_accuracy: 1.0000\n",
            "Epoch 22/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0145 - accuracy: 0.9815 - val_loss: 0.0020 - val_accuracy: 1.0000\n",
            "Epoch 23/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0149 - accuracy: 0.9815 - val_loss: 0.0015 - val_accuracy: 1.0000\n",
            "Epoch 24/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0140 - accuracy: 0.9815 - val_loss: 0.0011 - val_accuracy: 1.0000\n",
            "Epoch 25/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0152 - accuracy: 0.9815 - val_loss: 8.5356e-04 - val_accuracy: 1.0000\n",
            "Epoch 26/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0131 - accuracy: 0.9815 - val_loss: 0.0010 - val_accuracy: 1.0000\n",
            "Epoch 27/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0126 - accuracy: 0.9815 - val_loss: 8.0711e-04 - val_accuracy: 1.0000\n",
            "Epoch 28/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0113 - accuracy: 0.9815 - val_loss: 8.0723e-04 - val_accuracy: 1.0000\n",
            "Epoch 29/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0103 - accuracy: 0.9815 - val_loss: 9.3155e-04 - val_accuracy: 1.0000\n",
            "Epoch 30/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0096 - accuracy: 0.9815 - val_loss: 6.7605e-04 - val_accuracy: 1.0000\n",
            "Epoch 31/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0074 - accuracy: 0.9815 - val_loss: 7.4338e-04 - val_accuracy: 1.0000\n",
            "Epoch 32/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0053 - accuracy: 0.9815 - val_loss: 5.1661e-04 - val_accuracy: 1.0000\n",
            "Epoch 33/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0066 - accuracy: 0.9815 - val_loss: 4.3196e-04 - val_accuracy: 1.0000\n",
            "Epoch 34/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0032 - accuracy: 1.0000 - val_loss: 4.1503e-04 - val_accuracy: 1.0000\n",
            "Epoch 35/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0019 - accuracy: 1.0000 - val_loss: 5.0273e-04 - val_accuracy: 1.0000\n",
            "Epoch 36/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0022 - accuracy: 1.0000 - val_loss: 2.7973e-04 - val_accuracy: 1.0000\n",
            "Epoch 37/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0019 - accuracy: 1.0000 - val_loss: 2.5671e-04 - val_accuracy: 1.0000\n",
            "Epoch 38/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 0.0014 - accuracy: 1.0000 - val_loss: 2.4832e-04 - val_accuracy: 1.0000\n",
            "Epoch 39/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 9.5433e-04 - accuracy: 1.0000 - val_loss: 1.9596e-04 - val_accuracy: 1.0000\n",
            "Epoch 40/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 8.2636e-04 - accuracy: 1.0000 - val_loss: 1.7126e-04 - val_accuracy: 1.0000\n",
            "Epoch 41/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 6.8318e-04 - accuracy: 1.0000 - val_loss: 1.4628e-04 - val_accuracy: 1.0000\n",
            "Epoch 42/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.8021e-04 - accuracy: 1.0000 - val_loss: 8.7686e-05 - val_accuracy: 1.0000\n",
            "Epoch 43/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 7.8854e-04 - accuracy: 1.0000 - val_loss: 1.2050e-04 - val_accuracy: 1.0000\n",
            "Epoch 44/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.9542e-04 - accuracy: 1.0000 - val_loss: 6.8272e-05 - val_accuracy: 1.0000\n",
            "Epoch 45/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.7997e-04 - accuracy: 1.0000 - val_loss: 6.0790e-05 - val_accuracy: 1.0000\n",
            "Epoch 46/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.6622e-04 - accuracy: 1.0000 - val_loss: 4.0808e-05 - val_accuracy: 1.0000\n",
            "Epoch 47/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.8007e-04 - accuracy: 1.0000 - val_loss: 5.9728e-05 - val_accuracy: 1.0000\n",
            "Epoch 48/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.3778e-04 - accuracy: 1.0000 - val_loss: 4.8742e-05 - val_accuracy: 1.0000\n",
            "Epoch 49/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.9655e-04 - accuracy: 1.0000 - val_loss: 5.8357e-05 - val_accuracy: 1.0000\n",
            "Epoch 50/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.4269e-04 - accuracy: 1.0000 - val_loss: 2.9298e-05 - val_accuracy: 1.0000\n",
            "Epoch 51/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.7855e-04 - accuracy: 1.0000 - val_loss: 2.7998e-05 - val_accuracy: 1.0000\n",
            "Epoch 52/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.8146e-04 - accuracy: 1.0000 - val_loss: 2.4423e-05 - val_accuracy: 1.0000\n",
            "Epoch 53/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.8312e-04 - accuracy: 1.0000 - val_loss: 1.0635e-05 - val_accuracy: 1.0000\n",
            "Epoch 54/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 8.0707e-05 - accuracy: 1.0000 - val_loss: 2.3093e-05 - val_accuracy: 1.0000\n",
            "Epoch 55/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 8.3213e-05 - accuracy: 1.0000 - val_loss: 2.4565e-05 - val_accuracy: 1.0000\n",
            "Epoch 56/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 9.7983e-05 - accuracy: 1.0000 - val_loss: 8.5205e-06 - val_accuracy: 1.0000\n",
            "Epoch 57/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 6.9688e-05 - accuracy: 1.0000 - val_loss: 1.2395e-05 - val_accuracy: 1.0000\n",
            "Epoch 58/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 6.0192e-05 - accuracy: 1.0000 - val_loss: 8.0947e-06 - val_accuracy: 1.0000\n",
            "Epoch 59/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.9787e-05 - accuracy: 1.0000 - val_loss: 8.9048e-06 - val_accuracy: 1.0000\n",
            "Epoch 60/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.0237e-05 - accuracy: 1.0000 - val_loss: 5.1993e-06 - val_accuracy: 1.0000\n",
            "Epoch 61/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.4682e-05 - accuracy: 1.0000 - val_loss: 2.9481e-06 - val_accuracy: 1.0000\n",
            "Epoch 62/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.3122e-05 - accuracy: 1.0000 - val_loss: 5.4023e-06 - val_accuracy: 1.0000\n",
            "Epoch 63/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.1376e-05 - accuracy: 1.0000 - val_loss: 2.6491e-06 - val_accuracy: 1.0000\n",
            "Epoch 64/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.0058e-05 - accuracy: 1.0000 - val_loss: 2.2305e-06 - val_accuracy: 1.0000\n",
            "Epoch 65/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.0931e-05 - accuracy: 1.0000 - val_loss: 1.7392e-06 - val_accuracy: 1.0000\n",
            "Epoch 66/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.1944e-05 - accuracy: 1.0000 - val_loss: 3.1163e-06 - val_accuracy: 1.0000\n",
            "Epoch 67/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 9.4372e-06 - accuracy: 1.0000 - val_loss: 3.8209e-06 - val_accuracy: 1.0000\n",
            "Epoch 68/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.1170e-05 - accuracy: 1.0000 - val_loss: 3.1627e-06 - val_accuracy: 1.0000\n",
            "Epoch 69/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.3851e-05 - accuracy: 1.0000 - val_loss: 1.1915e-06 - val_accuracy: 1.0000\n",
            "Epoch 70/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.8748e-06 - accuracy: 1.0000 - val_loss: 8.2094e-07 - val_accuracy: 1.0000\n",
            "Epoch 71/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.6384e-06 - accuracy: 1.0000 - val_loss: 1.1043e-06 - val_accuracy: 1.0000\n",
            "Epoch 72/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.0320e-06 - accuracy: 1.0000 - val_loss: 5.4360e-07 - val_accuracy: 1.0000\n",
            "Epoch 73/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.3686e-06 - accuracy: 1.0000 - val_loss: 4.5273e-07 - val_accuracy: 1.0000\n",
            "Epoch 74/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.4776e-06 - accuracy: 1.0000 - val_loss: 5.9274e-07 - val_accuracy: 1.0000\n",
            "Epoch 75/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.6867e-06 - accuracy: 1.0000 - val_loss: 7.4219e-07 - val_accuracy: 1.0000\n",
            "Epoch 76/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.3203e-06 - accuracy: 1.0000 - val_loss: 2.7098e-07 - val_accuracy: 1.0000\n",
            "Epoch 77/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.7916e-06 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000\n",
            "Epoch 78/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.6606e-06 - accuracy: 1.0000 - val_loss: 1.7790e-07 - val_accuracy: 1.0000\n",
            "Epoch 79/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 9.9390e-07 - accuracy: 1.0000 - val_loss: 2.8055e-07 - val_accuracy: 1.0000\n",
            "Epoch 80/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.2509e-06 - accuracy: 1.0000 - val_loss: 3.8906e-07 - val_accuracy: 1.0000\n",
            "Epoch 81/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.4909e-06 - accuracy: 1.0000 - val_loss: 1.6870e-07 - val_accuracy: 1.0000\n",
            "Epoch 82/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.3822e-06 - accuracy: 1.0000 - val_loss: 1.2902e-07 - val_accuracy: 1.0000\n",
            "Epoch 83/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.2614e-07 - accuracy: 1.0000 - val_loss: 8.6862e-08 - val_accuracy: 1.0000\n",
            "Epoch 84/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 6.7414e-07 - accuracy: 1.0000 - val_loss: 1.6815e-07 - val_accuracy: 1.0000\n",
            "Epoch 85/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 9.6195e-07 - accuracy: 1.0000 - val_loss: 1.3785e-07 - val_accuracy: 1.0000\n",
            "Epoch 86/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.0753e-07 - accuracy: 1.0000 - val_loss: 9.6765e-08 - val_accuracy: 1.0000\n",
            "Epoch 87/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.6608e-07 - accuracy: 1.0000 - val_loss: 1.4705e-07 - val_accuracy: 1.0000\n",
            "Epoch 88/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.7083e-07 - accuracy: 1.0000 - val_loss: 4.8111e-08 - val_accuracy: 1.0000\n",
            "Epoch 89/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 3.0698e-07 - accuracy: 1.0000 - val_loss: 9.9545e-08 - val_accuracy: 1.0000\n",
            "Epoch 90/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.1238e-07 - accuracy: 1.0000 - val_loss: 4.5547e-08 - val_accuracy: 1.0000\n",
            "Epoch 91/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 2.4261e-07 - accuracy: 1.0000 - val_loss: 3.7516e-08 - val_accuracy: 1.0000\n",
            "Epoch 92/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.2927e-07 - accuracy: 1.0000 - val_loss: 5.5818e-08 - val_accuracy: 1.0000\n",
            "Epoch 93/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.8235e-07 - accuracy: 1.0000 - val_loss: 3.5976e-08 - val_accuracy: 1.0000\n",
            "Epoch 94/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 8.3451e-08 - accuracy: 1.0000 - val_loss: 2.4264e-08 - val_accuracy: 1.0000\n",
            "Epoch 95/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 1.3426e-07 - accuracy: 1.0000 - val_loss: 3.4905e-08 - val_accuracy: 1.0000\n",
            "Epoch 96/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.4683e-08 - accuracy: 1.0000 - val_loss: 9.0396e-08 - val_accuracy: 1.0000\n",
            "Epoch 97/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 7.3820e-08 - accuracy: 1.0000 - val_loss: 4.9122e-08 - val_accuracy: 1.0000\n",
            "Epoch 98/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 7.8562e-08 - accuracy: 1.0000 - val_loss: 1.8508e-08 - val_accuracy: 1.0000\n",
            "Epoch 99/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 5.5725e-08 - accuracy: 1.0000 - val_loss: 2.2753e-08 - val_accuracy: 1.0000\n",
            "Epoch 100/100\n",
            "54/54 [==============================] - 0s 2ms/step - loss: 4.9527e-08 - accuracy: 1.0000 - val_loss: 1.9830e-08 - val_accuracy: 1.0000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "n6mh8cPIpDeJ",
        "outputId": "7422e05b-ba3e-448c-9f23-ed48f6626e62"
      },
      "source": [
        "# Visualize the change in accuracy of model over the training process\n",
        "metrics = history.history\n",
        "plt.plot(history.epoch, metrics['accuracy'], metrics['val_accuracy'])\n",
        "plt.legend(['accuracy','val_accuracy'])\n",
        "plt.show()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "ZlInFOyjvUlP",
        "outputId": "a60662a2-bf41-410e-d867-595936e4fc02"
      },
      "source": [
        "# Visualize the loss of the model over the training process\n",
        "metrics = history.history\n",
        "plt.plot(history.epoch, metrics['loss'], metrics['val_loss'])\n",
        "plt.legend(['loss','val_loss'])\n",
        "plt.show()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HICFh5qLvXQ7",
        "outputId": "03eb1090-7d7c-42c8-eb0c-0e6adaaff088"
      },
      "source": [
        "# Test model on the test data and measure accuracy\n",
        "predictions = np.argmax(model.predict(inputs_test), axis=1)\n",
        "true_values = np.argmax(outputs_test, axis=1)\n",
        "\n",
        "test_acc = sum(predictions == true_values) / len(true_values)\n",
        "print(f'Test set accuracy: {test_acc:.0%}')"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test set accuracy: 100%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 501
        },
        "id": "qreqMKxQwDXm",
        "outputId": "1c54f9f4-9171-4985-e19e-522ad8fc6b85"
      },
      "source": [
        "# Plot a confusion matrix for model on test data\n",
        "confusion_mtx = tf.math.confusion_matrix(true_values, predictions) \n",
        "plt.figure(figsize=(10, 8))\n",
        "sns.heatmap(confusion_mtx, xticklabels=ACTIONS, yticklabels=ACTIONS, \n",
        "            annot=True, fmt='g')\n",
        "plt.xlabel('Prediction')\n",
        "plt.ylabel('Label')\n",
        "plt.show()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MsyXiukYxhqY",
        "outputId": "7e80a101-7dc8-4a22-b22d-d8d1ce7b8b33"
      },
      "source": [
        "# Convert the model to the TensorFlow Lite format without quantization\n",
        "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
        "tflite_model = converter.convert()\n",
        "\n",
        "# Save the model to disk\n",
        "open(\"gesture_model.tflite\", \"wb\").write(tflite_model)\n",
        "  \n",
        "import os\n",
        "basic_model_size = os.path.getsize(\"gesture_model.tflite\")\n",
        "print(\"Model is %d bytes\" % basic_model_size)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: /tmp/tmpjrkiz_cs/assets\n",
            "Model is 65128 bytes\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Mp8J6BHmyK_I",
        "outputId": "df9fb315-f45a-4338-f06a-42947b3a1488"
      },
      "source": [
        "!apt-get -qq install xxd"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Selecting previously unselected package xxd.\n",
            "(Reading database ... 160772 files and directories currently installed.)\n",
            "Preparing to unpack .../xxd_2%3a8.0.1453-1ubuntu1.4_amd64.deb ...\n",
            "Unpacking xxd (2:8.0.1453-1ubuntu1.4) ...\n",
            "Setting up xxd (2:8.0.1453-1ubuntu1.4) ...\n",
            "Processing triggers for man-db (2.8.3-2ubuntu0.1) ...\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oPJhsnglyMhO",
        "outputId": "85c2c3e0-be69-44a3-d44b-fb7b949530d6"
      },
      "source": [
        "!echo \"const unsigned char model[] = {\" > /content/model.h\n",
        "!cat gesture_model.tflite | xxd -i      >> /content/model.h\n",
        "!echo \"};\"                              >> /content/model.h\n",
        "\n",
        "import os\n",
        "model_h_size = os.path.getsize(\"model.h\")\n",
        "print(f\"Header file, model.h, is {model_h_size:,} bytes.\")\n",
        "print(\"\\nOpen the side panel (refresh if needed). Double click model.h to download the file.\")"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Header file, model.h, is 401,658 bytes.\n",
            "\n",
            "Open the side panel (refresh if needed). Double click model.h to download the file.\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}