no code implementations • ICLR 2019 • Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo
To solve this, we propose a concept to learn a mapping that embeds both images and attributes to the shared representation space that can be generalized even for unseen classes by interpolating from the information of seen classes, which we refer to shared manifold learning.
no code implementations • 14 Apr 2024 • Kota Tanabe, Masahiro Suzuki, Hiroki Sakaji, Itsuki Noda
To achieve this, we propose an instruction tuning data in Japanese called JaFIn, the Japanese Financial Instruction Dataset.
1 code implementation • 12 Mar 2024 • Yuta Oshima, Shohei Taniguchi, Masahiro Suzuki, Yutaka Matsuo
In the experiments, we first evaluate our SSM-based model with UCF101, a standard benchmark of video generation.
no code implementations • 22 Feb 2024 • Takehiro Takayanagi, Masahiro Suzuki, Ryotaro Kobayashi, Hiroki Sakaji, Kiyoshi Izumi
Causality is fundamental in human cognition and has drawn attention in diverse research fields.
no code implementations • 19 Oct 2023 • Masahiro Suzuki, Shomu Furuta, Yusuke Fukazawa
We adopted a personalized model to predict the individual's movement trajectory from their data, instead of predicting from the overall movement, based on the hypothesis that human movement is unique to each person.
no code implementations • 16 Oct 2023 • Issey Sukeda, Masahiro Suzuki, Hiroki Sakaji, Satoshi Kodera
Furthermore, our results underscore the potential of adapting English-centric models for Japanese applications in domain adaptation, while also highlighting the persisting limitations of Japanese-centric models.
1 code implementation • 7 Sep 2023 • Masahiro Suzuki, Masanori Hirano, Hiroki Sakaji
We performed Low-Rank Adaptation (LoRA) tuning on both Japanese and English existing models using our instruction dataset.
1 code implementation • 31 May 2023 • Shohei Taniguchi, Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo
We address the problem of biased gradient estimation in deep Boltzmann machines (DBMs).
1 code implementation • 22 May 2023 • Masanori Hirano, Masahiro Suzuki, Hiroki Sakaji
There are two ways to support languages other than English by those LLMs: constructing LLMs from scratch or tuning existing models.
no code implementations • 14 Jan 2023 • Tadahiro Taniguchi, Shingo Murata, Masahiro Suzuki, Dimitri Ognibene, Pablo Lanillos, Emre Ugur, Lorenzo Jamone, Tomoaki Nakamura, Alejandra Ciria, Bruno Lara, Giovanni Pezzulo
Therefore, in this paper, we clarify the definitions, relationships, and status of current research on these topics, as well as missing pieces of world models and predictive coding in conjunction with crucially related concepts such as the free-energy principle and active inference in the context of cognitive and developmental robotics.
no code implementations • 5 Jul 2022 • Masahiro Suzuki, Yutaka Matsuo
In recent years, deep generative models, i. e., generative models in which distributions are parameterized by deep neural networks, have attracted much attention, especially variational autoencoders, which are suitable for accomplishing the above challenges because they can consider heterogeneity and infer good representations of data.
1 code implementation • 1 Dec 2021 • Masahiro Suzuki
In this paper, we explore the tokenized representation of musical scores using the Transformer model to automatically generate musical scores.
no code implementations • 13 Oct 2021 • Kazutoshi Shinoda, Yuki Takezawa, Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo
An interactive instruction following task has been proposed as a benchmark for learning to map natural language instructions and first-person vision into sequences of actions to interact with objects in 3D environments.
no code implementations • 29 Sep 2021 • Masahiro Suzuki, Yutaka Matsuo
A state-of-the-art approach to learning this aggregation of experts is to encourage all modalities to be reconstructed and cross-generated from arbitrary subsets.
no code implementations • 29 Sep 2021 • Yuya Kobayashi, Masahiro Suzuki, Yutaka Matsuo
Therefore, we introduce several crucial components which help inference and training with the proposed model.
no code implementations • 28 Jul 2021 • Masahiro Suzuki, Takaaki Kaneko, Yutaka Matsuo
With the recent rapid progress in the study of deep generative models (DGMs), there is a need for a framework that can implement them in a simple and generic way.
no code implementations • 15 Mar 2021 • Tadahiro Taniguchi, Hiroshi Yamakawa, Takayuki Nagai, Kenji Doya, Masamichi Sakagami, Masahiro Suzuki, Tomoaki Nakamura, Akira Taniguchi
This approach is based on two ideas: (1) brain-inspired AI, learning human brain architecture to build human-level intelligence, and (2) a probabilistic generative model(PGM)-based cognitive system to develop a cognitive system for developmental robots by integrating PGMs.
no code implementations • 1 Jan 2021 • Hitoshi Nakanishi, Masahiro Suzuki, Yutaka Matsuo
Moreover, there is objective mismatching that models are trained to minimize total reconstruction errors while we expect a small deviation on normal pixels and large deviation on anomalous pixels.
no code implementations • ICLR 2020 • Hirono Okamoto, Masahiro Suzuki, Yutaka Matsuo
However, on difficult datasets or models with low classification ability, these methods incorrectly regard in-distribution samples close to the decision boundary as OOD samples.
no code implementations • 20 Oct 2019 • Tadahiro Taniguchi, Tomoaki Nakamura, Masahiro Suzuki, Ryo Kuniyasu, Kaede Hayashi, Akira Taniguchi, Takato Horii, Takayuki Nagai
The model is called VAE+GMM+LDA+ASR.
Automatic Speech Recognition Automatic Speech Recognition (ASR) +1
no code implementations • 25 Sep 2019 • Masahiro Suzuki, Yutaka Matsuo
However, this relation-based approach presents a difficulty: many of the test images are predicted as biased to the seen domain, i. e., the \emph{domain bias problem}.
no code implementations • ICLR 2019 • Hirono Okamoto, Shohei Ohsawa, Itto Higuchi, Haruka Murakami, Mizuki Sango, Zhenghang Cui, Masahiro Suzuki, Hiroshi Kajino, Yutaka Matsuo
It reformulates the posterior with a natural paring $\langle, \rangle: \mathcal{Z} \times \mathcal{Z}^* \rightarrow \Real$, which can be expanded to uncountable infinite domains such as continuous domains as well as interpolation.
no code implementations • ICLR Workshop DeepGenStruct 2019 • Hirono Okamoto, Masahiro Suzuki, Itto Higuchi, Shohei Ohsawa, Yutaka Matsuo
However, when the dimension of multiclass labels is large, these models cannot change images corresponding to labels, because learning multiple distributions of the corresponding class is necessary to transfer an image.
no code implementations • 26 Jan 2018 • Masahiro Suzuki, Kotaro Nakayama, Yutaka Matsuo
However, we found that when this model attempts to generate a large dimensional modality missing at the input, the joint representation collapses and this modality cannot be generated successfully.
no code implementations • 25 Nov 2016 • Joji Toyama, Masanori Misono, Masahiro Suzuki, Kotaro Nakayama, Yutaka Matsuo
The report of earlier studies has introduced a latent variable to capture the entire meaning of sentence and achieved improvement on attention-based Neural Machine Translation.
2 code implementations • 7 Nov 2016 • Masahiro Suzuki, Kotaro Nakayama, Yutaka Matsuo
As described herein, we propose a joint multimodal variational autoencoder (JMVAE), in which all modalities are independently conditioned on joint representation.
no code implementations • 10 Oct 2016 • Masatoshi Uehara, Issei Sato, Masahiro Suzuki, Kotaro Nakayama, Yutaka Matsuo
Generative adversarial networks (GANs) are successful deep generative models.