HUMAN FACES GENERATION USING DCGAN || VIP || IEEE || HYD || PROJECTS ||

TO PURCHASE OUR PROJECTS IN ONLINE (OR) OFFLINE
CONTACT:VENKAT INNOVATIVE PROJECTS
NAME: VENKATARAO GANIPISETTY
Mobile & WhatsApp :+91 9966499110
Mobile & WhatsApp :+91 9573201550
Email :[email protected]
Email :[email protected]
website:https://venkatinnovativeprojects.com/
ABOUT PROJECT:

In propose project we are employing DCGAN (deep convolution generative adversarial network) to generate human faces. DCGAN consists of two different networks such as Generator and Discriminator where generator will generate faces from trained model and then Discriminator will predict weather uploaded and generated faces are real or fake.
Text to Image Generation
In this module we have used combination of CNN and LSTM algorithm to generate images from text where CNN will be responsible to generate images by utilizing text features predicted by LSTM. This model trained on Image Net dataset and this model will be applied on LSTM features to generated related image. In real time they will train on millions of images so generation will be accurate but we trained the model on 8000 images only so it will work for few sentences which we have given in ‘samples_input_text.txt’ file.
To implement this project we have designed following modules
1) User Login: user can login to system using username and password as ‘admin and admin’.
2) Train & Load DCGAN: after login user will use this module to generate and load DCGAN model and then perform prediction on test data to calculate prediction accuracy
3) Human Faces Generation: using this module user can upload any image and then DCGAN will generate some images and then apply Discriminator to detect those images as Real or Fake (Ai generated).
4) Text to Image: in this module user can upload some text and then application will generate related image.