assets/hub/pytorch_vision_googlenet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### This notebook is optionally accelerated with a GPU runtime.\n",
    "### If you would like to use this acceleration, please select the menu option \"Runtime\" -> \"Change runtime type\", select \"Hardware Accelerator\" -> \"GPU\" and click \"SAVE\"\n",
    "\n",
    "----------------------------------------------------------------------\n",
    "\n",
    "# GoogLeNet\n",
    "\n",
    "*Author: Pytorch Team*\n",
    "\n",
    "**GoogLeNet was based on a deep convolutional neural network architecture codenamed \"Inception\" which won ImageNet 2014.**\n",
    "\n",
    "_ | _\n",
    "- | -\n",
    "![alt](https://pytorch.org/assets/images/googlenet1.png) | ![alt](https://pytorch.org/assets/images/googlenet2.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "model = torch.hub.load('pytorch/vision', 'googlenet', pretrained=True)\n",
    "model.eval()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All pre-trained models expect input images normalized in the same way,\n",
    "i.e. mini-batches of 3-channel RGB images of shape `(3 x H x W)`, where `H` and `W` are expected to be at least `224`.\n",
    "The images have to be loaded in to a range of `[0, 1]` and then normalized using `mean = [0.485, 0.456, 0.406]`\n",
    "and `std = [0.229, 0.224, 0.225]`.\n",
    "\n",
    "Here's a sample execution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Download an example image from the pytorch website\n",
    "import urllib\n",
    "url, filename = (\"https://github.com/pytorch/hub/raw/master/dog.jpg\", \"dog.jpg\")\n",
    "try: urllib.URLopener().retrieve(url, filename)\n",
    "except: urllib.request.urlretrieve(url, filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# sample execution (requires torchvision)\n",
    "from PIL import Image\n",
    "from torchvision import transforms\n",
    "input_image = Image.open(filename)\n",
    "preprocess = transforms.Compose([\n",
    "    transforms.Resize(256),\n",
    "    transforms.CenterCrop(224),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),\n",
    "])\n",
    "input_tensor = preprocess(input_image)\n",
    "input_batch = input_tensor.unsqueeze(0) # create a mini-batch as expected by the model\n",
    "\n",
    "# move the input and model to GPU for speed if available\n",
    "if torch.cuda.is_available():\n",
    "    input_batch = input_batch.to('cuda')\n",
    "    model.to('cuda')\n",
    "\n",
    "with torch.no_grad():\n",
    "    output = model(input_batch)\n",
    "# Tensor of shape 1000, with confidence scores over Imagenet's 1000 classes\n",
    "print(output[0])\n",
    "# The output has unnormalized scores. To get probabilities, you can run a softmax on it.\n",
    "print(torch.nn.functional.softmax(output[0], dim=0))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Description\n",
    "\n",
    "GoogLeNet was based on a deep convolutional neural network architecture codenamed \"Inception\", which was responsible for setting the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC 2014). The 1-crop error rates on the ImageNet dataset with a pretrained model are list below.\n",
    "\n",
    "| Model structure | Top-1 error | Top-5 error |\n",
    "| --------------- | ----------- | ----------- |\n",
    "|  googlenet       | 30.22       | 10.47       |\n",
    "\n",
    "\n",
    "\n",
    "### References\n",
    "\n",
    " - [Going Deeper with Convolutions](https://arxiv.org/abs/1409.4842)"
   ]
  }
 ],
 "metadata": {},
 "nbformat": 4,
 "nbformat_minor": 2
}