Word Embeddings for Speech Recognition

   Abstract