The Straight-Through Estimator (STE) is widely used for back-propagating gradients through the quantization function, but the STE technique lacks a complete theoretical understanding. We propose an alternative methodology…

Recurrent neural networks (RNNs) have shown state of the art results for speech recognition, natural language processing, image captioning and video summarizing applications. Many of these applications…

Machine learning-based applications are increasingly prevalent in IoT devices. The power and storage constraints of these devices make it particularly challenging to run modern neural networks, limiting the…