Quick and Simple Image Recognition with Inception v3

• Based on Google's Image Recognition Tutorial
• Detailed Example: KerasInception
• A Web based Node.js API
• Python Example on Heroku
• Class Labels ImageNet

General Fast Git

In a Terminal:

git clone https://github.com/EN10/SimpleInception.git
cd SimpleInception
bash install.sh

Colab Version

!wget https://raw.githubusercontent.com/tensorflow/models/master/tutorials/image/imagenet/classify_image.py
!wget https://raw.githubusercontent.com/EN10/SimpleInception/master/5918348-image.jpg
!python classify_image.py --image_file 5918348-image.jpg

Fast Cloud9 Version

"Go To Your Dashboard"
Click: "+ Create a new workspace"
Give a "Workspace name" e.g. inception
"Clone from Git URL":

https://github.com/EN10/SimpleInception.git

"Choose a template": Blank
Click "Create workspace" (~35s)

then "+ New Terminal" type:

bash install.sh

Individual Steps:

Get Program:

wget https://raw.githubusercontent.com/tensorflow/models/master/tutorials/image/imagenet/classify_image.py

Updated 10/12/16

Get Image:

wget https://raw.githubusercontent.com/EN10/SimpleInception/master/5918348-image.jpg

Install Tensorflow:

sudo pip install -U pip
sudo pip install tensorflow

Run Prediction:

sudo python classify_image.py --image_file 5918348-image.jpg

Runs in < 14s First Time!

ERROR:

Disable Error Messages:

export TF_CPP_MIN_LOG_LEVEL=2