Eli Sheppard

I am using deep LSTMs to automatically recognise speech. My hope is to produce a system which will be naturally integratible with neural network based image recognition to allow for speech grounding to be performed by combining speech and visual signals.