readme.md

image2vec

1g_sorted.py

• f8k, f30k
• environment: python2
• description: sort the captions and token

2g_pair.py

• f8k, f30k
• python3
• pairs of images and captions

3g_pkl.py

• common
• python2
• Generate the pickle file of 1b_corpus dataset

4g_model.py

• common
• python2
• Generate word2vec model of 1b_corpus

5g_filtered_pair.py

• f8k, f30k
• python2
• Generate filtered pair of file name and vector

6g_image_pair.py

• f8k, f30k
• python2
• Generate pair of image data and caption

7cnn.py

• f8k, f30k
• python2
• train the model with cnn