Dataset is street snap images(128x128).

This network is generate images of 128x128 pixels.
These are generated images by DCGAN.

Use Pre-trained Model : GoogleNews-vectors-negative300

• translator(Ko => En) -> Word2Vec(Pre-trained model)
  • similarity(input_text, gender_words, category_words)
    • Find Best Match Image => Generate Image.
  • Image Analysis(Gender, Category)
• DCGAN -> Discriminator Model(Good / Bad)