Lookahead Convolution Layer for Unidirectional Recurrent Neural Networks

By Chongshun Wang, Dani Yogatama, Adam Coates, T. Han, Awni Y. Hannun, Bo Xiao. 2016

15 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

TED-LIUM 3: twice as much data and corpus repartition for experiments on speaker adaptation (2018, background): “This architecture is composed of nc convolution layers (CNN), followed by nr uni or bidirectional recurrent layers, a lookahead convolution layer [15], and one fully connected layer just before the softmax layer, as shown in Figure 1.”
Online and Linear-Time Attention by Enforcing Monotonic Alignments (2017, neutral): “While these figures show a clear advantage to our approach, our model architecture differed significantly from those of (Luo et al., 2016; Wang et al., 2016).”
Monotonic Chunkwise Attention (2017, background): “Under review as a conference paper at ICLR 2018 Prior Result WER (Raffel et al., 2017) (CTC baseline) 33.4% (Luo et al., 2016) (Reinforcement Learning) 27.0% (Wang et al., 2016) (CTC) 22.7% (Raffel et al., 2017) Table 2: ROUGE F-scores for document summarization on the CNN/Daily Mail dataset.”