What is Deep Trans?

DeepTrans is a character level language model for transliterating English text into Hindi. It is based on the attention mechanism presented in [1] and its implementation in Tensorflow’s Sequence to Sequence models. This project has been inspired by the translation model presented in tensorflow’s sequence to sequence model. This project comes with a pretrained model (2 layers with 256 units each) for Hindi but can be easily trained over the existing model or from scratch. The pretrained models are trained on lowercase words. If you wish to train your own model then feel free to do whatever you want and we would be glad if you could share your results and models with me. we hope to see interesting results.

Code and references

The code is public and would like to have more people collaborate with us to make it better.

Project Details

Date: Oct 10, 2016

Author: Ayushman Dash

Categories: projectseq2seq


Related Works.

Bidirectional Learning for Robust Neural Networks

Recognizing Challenging Handwritten Annotations with Fully Convolutional Networks

Real-Time Document Image Classification using Deep CNN and Extreme Learning Machines

Cutting the Error by Half: Investigation of Very Deep CNN and Advanced Training Strategies for Document Image Classification

TAC-GAN - Text Conditioned Auxiliary Classifier Generative Adversarial Network


We believe that creativity comes with freedom and to solve challenging problems in AI we need to have freedom and creative thinking. We provide an unconstrained environment to highly motivated students to do whatever that comes to their minds and explore deep learning.

Social Links

Our Bunker

Building 86,
Davenport 11,
67663 Kaiserslautern,