amznbin

Project Description

amznbin is a deep learning image recognition project using Amazon Bin Image Dataset.

We solve two tasks,

1. Count: Predict the number of items in the image
2. Classify: Predict what item is contained in the image

To reduce the complexity of the tasks and training time,

1. We use images with only one kind of item. - this enables us to solve classify task as a single classification task
2. We specify the minimum number of repetition of an item - most items appear only once in all images, when untreated it will make testing and validating unconvincing.
3. We manually ruled out a number of invalid images. - check dataset/invalid_images.json to see which images are excluded.

How to Start

• On root directory run pip install -r requirements.txt to install required libraries
• Run make image_dataset and make load_metadata on command line to download the images and metadata (Using tmux would be a good idea)
• After make load_metadata is finished, run make prepare_train and input the number of repetition you want (we recommend 20 or higher) and data augmentation mode.
• When all above is finished you are ready to run our program!

Demo

• Run make demo to checkout our training demo  - To run the demo of our pretrained model, open jupyter notebook and run src/demo.ipynb
  • our pretrained model was too big to upload on github, you can train the model and use your own trained model to use this file

Results

We accomplished 68.5714% accuracy on count task, and 41.4286% accuracy on classify task