Kenji's Home Page

Kenji Imamura (今村 賢治)

A researcher of natural language processing.
My research interests include: 1) machine translation, 2) NLP using neural networks, 3) natural language generation, 4) natural language analysis (dependency parsing, predicate-argument structure analysis).

Recent Publications

2023

NAIST-NICT WMT’23 General MT Task Submission

Pivot Translation for Zero-resource Language Pairs Based on a Multilingual Pretrained Model

A Study of Pivot Translation between Non-English Languages Using Multilingual Models

Subword Addtion to a SentencePiece Tokenizer for Multilingual Pretrained Models

Construction of GCP Simultaneous Corpus

2022

NAIST-NICT-TIT WMT22 General MT Task Submission

Extending the Subwording Model of Multilingual Pretrained Models for New Languages

Speeding up of Multilingual Pretrained Machine Translation by Vocabulary Selection

A Survey of Machine Translation Technologies for Machine Translation on Patent Documents

2021

Language Expansion of a Pretrained Model and its Effects

NICT-2 Translation System at WAT-2021: Applying a Pretrained Multilingual Encoder-Decoder Model to Low-resource Language Pairs

A Study of Speed-up in Neural Machine Translation Combined with Knowledge Distillation.

2020

Transformer-based Double-token Bidirectional Autoregressive Decoding in Neural Machine Translation

Improvement of Neural Machine Translation that Uses Non-autoregressive Decoding

Neural Machine Translation Using Multiple Back-translation Generated by Sampling

Language Translation

2019

Recycling a Pre-trained BERT Encoder for Neural Machine Translation

Long Warm-up and Self-Training: Training Strategies of NICT-2 NMT System at WAT-2019

Neural Machine Translation that Recycles Pre-trained BERT Encoder

Exploiting Out-of-Domain Parallel Data through Multilingual Transfer Learning for Low-Resource Neural Machine Translation

Usage of Long Context in Dialogue Translation

Recent Pictures

StartBucks

Post Office

Bank

Near Ise-Jingu Shrine, Aug. 2024

Deerlet 2024
May, 2024