111 citations · 241 across the 26 of their papers we have counts for
9 papers · 2 filters
Another Diversity-Promoting Objective Function for Neural Dialogue Generation
Ryo Nakamura, Katsuhito Sudoh, Koichiro Yoshino +1
Although generation-based dialogue systems have been widely researched, the response generations by most existing systems have very low diversities. The most likely reason for this…
Multi-Source Neural Machine Translation with Data Augmentation
Yuta Nishimura, Katsuhito Sudoh, Graham Neubig +1
Multi-source translation systems translate from multiple languages to a single target language. By using information from these multiple sources, these systems achieve large gains…
End-to-End Feedback Loss in Speech Chain Framework via Straight-Through Estimator
Andros Tjandra, Sakriani Sakti, Satoshi Nakamura
The speech chain mechanism integrates automatic speech recognition (ASR) and text-to-speech synthesis (TTS) modules into a single cycle during training. In our previous work, we ap…
Training Neural Machine Translation using Word Embedding-based Loss
Katsuki Chousa, Katsuhito Sudoh, Satoshi Nakamura
In neural machine translation (NMT), the computational cost at the output layer increases with the size of the target-side vocabulary. Using a limited-size vocabulary instead may c…
Multi-scale Alignment and Contextual History for Attention Mechanism in Sequence-to-sequence Model
Andros Tjandra, Sakriani Sakti, Satoshi Nakamura
A sequence-to-sequence model is a neural network module for mapping two sequences of different lengths. The sequence-to-sequence model has three core modules: encoder, decoder, and…
Multi-Source Neural Machine Translation with Missing Data
Yuta Nishimura, Katsuhito Sudoh, Graham Neubig +1
Multi-source translation is an approach to exploit multiple inputs (e.g. in two different languages) to increase translation accuracy. In this paper, we examine approaches for mult…