no code implementations • 22 Apr 2021 • Janne Pylkkönen, Antti Ukkonen, Juho Kilpikoski, Samu Tamminen, Hannes Heikinheimo
In this paper we show that RNN-transducer models can be effectively adapted to new domains using only small amounts of textual data.
no code implementations • 16 Jun 2020 • Nikolaj Tatti, Hannes Heikinheimo
Such itemset families define a probabilistic model for the data from which the original collection of itemsets has been derived from.
1 code implementation • 10 Dec 2016 • Matti Lankinen, Hannes Heikinheimo, Pyry Takala, Tapani Raiko, Juha Karhunen
Inspired by recent research, we explore ways to model the highly morphological Finnish language at the level of characters while maintaining the performance of word-level models.