Papers - SUGIURA, Komei

配置物体を考慮したAttention機構マルチモーダルネットワークによる衝突予測

河合竹彦, 平川翼, 山下隆義, 藤吉弘亘, 杉浦孔明

日本ロボット学会学術講演会予稿集(CD-ROM) 39th 2021

生活支援ロボットによる物体配置タスクにおける危険性予測および視覚的説明生成

畑中駿平, 上田雄斗, 植田有咲, 平川翼, 山下隆義, 藤吉弘亘, 杉浦孔明

日本ロボット学会学術講演会予稿集(CD-ROM) 39th 2021

物体指示理解タスクにおけるクロスモーダル言語生成に基づくデータ拡張

飯田紡, 九曜克之, 石川慎太朗, 杉浦孔明

日本ロボット学会学術講演会予稿集(CD-ROM) 39th 2021

深層学習を用いた太陽フレア予測モデルによる予報運用評価

西塚直人, 久保勇樹, 杉浦孔明, 田光江, 石井守

日本天文学会年会講演予稿集 2021 2021

ISSN  1347-0639

Visual Explanation using Attention Mechanism in Actor-Critic-based Deep Reinforcement Learning.

Hidenori Itaya, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi, Komei Sugiura

International Joint Conference on Neural Networks(IJCNN) (IEEE)     1 - 10 2021

• Access to Document (DOI)

Holding apparatus, container provided with tag, object holding program and object holding method

Y Onuma, K Sugiura

US Patent 11,192,242  2021

太陽フレア AI 予報 Deep Flare Net の運用化

西塚直人， 久保勇樹， 杉浦孔明， 田光江， 石井守

宇宙航空研究開発機構特別資料: 第 16 回 「宇宙環境シンポジウム」 講演論文集, 61-64  2020

Research paper (conference, symposium, etc.), Joint Work

Latent-Space Data Augmentation for Visually-Grounded Language Understanding

A Magassouba, K Sugiura, H Kawai

Advances in Artificial Intelligence. JSAI 2019. Advances in Intelligent Systems and Computing (Springer)  1128   179 - 187 2020

Research paper (other academic), Joint Work

• Access to Document (DOI)

A Multimodal Target-Source Classifier With Attention Branches to Understand Ambiguous Instructions for Fetching Daily Objects

A Magassouba, K Sugiura, H Kawai

IEEE Robotics and Automation Letters 5 ( 2 ) 532 - 539 2020

Research paper (scientific journal), Joint Work, Accepted

• Access to Document (DOI)

Multimodal Attention Branch Network に基づく把持命令文の生成

小椋忠志， 杉浦孔明， 平川翼， 山下隆義， 藤吉弘亘， 河井恒

人工知能学会全国大会論文集 第 34 回全国大会 (2020), 1Q3GS1105-1Q3GS1105 (The Japanese Society for Artificial Intelligence)  2020 ( 0 ) 1Q3GS1105 - 1Q3GS1105 2020

　View Summary

<p>Domestic service robots (DSRs) are a promising solution to the shortage of home care workers. Nonetheless, one of the main limitations of DSRs is their inability to naturally interact through language. Recently, data-driven approaches have been shown to be effective for tackling this limitation, however, they often require large-scale datasets, which is costly. Based on this background, we aim to perform automatic sentence generation for fetching instructions, e.g., ``Bring me a green tea bottle on the table.'' This is particularly challenging because appropriate expressions depend on the target object, as well as its surroundings. In this paper, we propose a method that generates sentences from visual inputs. Unlike other approaches, the proposed method has multimodal attention branches that utilize subword-level attention and generate sentences based on subword embeddings. In the experiment, we compared the proposed method with a baseline method using four standard metrics in image captioning. Experimental results show that the proposed method outperformed the baseline in terms of these metrics.</p>

• Access to Document (DOI)

Compensation on x-vector for Short Utterance Spoken Language Identification}}

P Shen, X Lu, K Sugiura, S Li, H Kawai

Proc. Odyssey 2020 The Speaker and Language Recognition Workshop, 47-52  2020

Alleviating the Burden of Labeling: Sentence Generation by Attention Branch Encoder-Decoder Network

T Ogura, A Magassouba, K Sugiura, T Hirakawa, T Yamashita, ...

IEEE Robotics and Automation Letters abs/2007.04557 2020

A3C における Attention 機構を用いた視覚的説明

板谷英典， 平川翼， 山下隆義， 藤吉弘亘， 杉浦孔明

人工知能学会全国大会論文集 第 34 回全国大会 (2020), 2J6GS204-2J6GS204 (The Japanese Society for Artificial Intelligence)  2020 ( 0 ) 2J6GS204 - 2J6GS204 2020

　View Summary

<p>Asynchronous Advantage Actor-Critic (A3C) is a representative method of deep reinforcement learning and is possible to solve difficult tasks such as games and robot control. However, it is difficult for deep reinforcement learning including A3C to understand and to explain the reason of action selection. To address this problem, we propose a method called a Mask-Attention A3C, which performs mask processing on feature map of Policy branch using attention map. The propose method can obtain an attention map that is useful for a visual explanation of agent behavior. In the experiment with Atari2600, we compare the scores in each game and demonstrate that the attention map obtained from our method is useful for visual explanation. In addition, we evaluate the explainability of obtained attention map using the scores of each game by changing the attention region.</p>

• Access to Document (DOI)

Reliable Probability Forecast of Solar Flares: Deep Flare Net-Reliable (DeFN-R)

N Nishizuka, Y Kubo, K Sugiura, M Den, M Ishii

The Astrophysical Journal 899 (2), 150  2020

Compensation on x-vector for short utterance spoken language identification

P Shen, X Lu, K Sugiura, S Li, H Kawai

Proc. Odyssey 2020 The Speaker and Language Recognition Workshop, 47-52  2020

Alleviating the Burden of Labeling: Sentence Generation by Attention Branch Encoder–Decoder Network

T Ogura, A Magassouba, K Sugiura, T Hirakawa, T Yamashita, ...

IEEE Robotics and Automation Letters 5 (4), 5945-5952 (Institute of Electrical and Electronics Engineers (IEEE))  5 ( 4 ) 5945 - 5952 2020

Research paper (scientific journal), Joint Work, Accepted

• Access to Document (DOI)

Instruction understanding system and instruction understanding method

K Sugiura, H Kawai

US Patent 10,796,098  2020

Understanding Natural Language Instructions for Fetching Daily Objects Using GAN-Based Multimodal Target–Source Classification

A Magassouba, K Sugiura, AT Quoc, H Kawai

IEEE Robotics and Automation Letters 4 ( 4 ) 3884 - 3891 2019

Research paper (scientific journal), Joint Work, Accepted

Operation of Solar Flare Prediction by Deep Flare Net.

N Nishizuka, K Sugiura, Y Kubo, M Den, M Ishii

Geophysical Research Abstracts 21  2019

Multimodal attention branch network for perspective-free sentence generation

A Magassouba, K Sugiura, H Kawai

arXiv preprint arXiv:1909.05664 abs/1909.05664 2019