Zobrazeno 1 - 10
of 402
pro vyhledávání: '"RICE, DANIEL"'
Autor:
Wu, Shanshan, Dimakis, Alexandros G., Sanghavi, Sujay, Yu, Felix X., Holtmann-Rice, Daniel, Storcheus, Dmitry, Rostamizadeh, Afshin, Kumar, Sanjiv
Linear encoding of sparse vectors is widely popular, but is commonly data-independent -- missing any possible extra (but a priori unknown) structure beyond sparsity. In this paper we present a new method to learn linear encoders that adapt to data, w
Externí odkaz:
http://arxiv.org/abs/1806.10175
Autor:
Kumar, Shankar, Nirschl, Michael, Holtmann-Rice, Daniel, Liao, Hank, Suresh, Ananda Theertha, Yu, Felix
Publikováno v:
Proceedings of ASRU 2017
Recurrent neural network (RNN) language models (LMs) and Long Short Term Memory (LSTM) LMs, a variant of RNN LMs, have been shown to outperform traditional N-gram LMs on speech recognition tasks. However, these models are computationally more expensi
Externí odkaz:
http://arxiv.org/abs/1711.05448
We develop a linear algebraic framework for the shape-from-shading problem, because tensors arise when scalar (e.g. image) and vector (e.g. surface normal) fields are differentiated multiple times. The work is in two parts. In this first part we inve
Externí odkaz:
http://arxiv.org/abs/1705.05885
We develop a linear algebraic framework for the shape-from-shading problem, because tensors arise when scalar (e.g. image) and vector (e.g. surface normal) fields are differentiated multiple times. Using this framework, we first investigate when imag
Externí odkaz:
http://arxiv.org/abs/1705.05902