no code implementations • 27 May 2024 • Junyoung Seo, Kazumi Fukuda, Takashi Shibuya, Takuya Narihira, Naoki Murata, Shoukang Hu, Chieh-Hsin Lai, Seungryong Kim, Yuki Mitsufuji
In these methods, an input view is geometrically warped to novel views with estimated depth maps, then the warped image is inpainted by T2I models.
no code implementations • 28 Mar 2023 • Hiromichi Kamata, Yuiko Sakuma, Akio Hayakawa, Masato Ishii, Takuya Narihira
We propose a high-quality 3D-to-3D conversion method, Instruct 3D-to-3D.
no code implementations • 23 Mar 2023 • Yuiko Sakuma, Masato Ishii, Takuya Narihira
We address the challenge of training a large supernet for the object detection task, using a relatively small amount of training data.
1 code implementation • 2 Feb 2023 • Kazuki Yoshiyama, Takuya Narihira
The goal of inverse rendering is to decompose geometry, lights, and materials given pose multi-view images.
1 code implementation • 5 Dec 2022 • Naoki Matsunaga, Masato Ishii, Akio Hayakawa, Kenji Suzuki, Takuya Narihira
Our goal is to develop fine-grained real-image editing methods suitable for real-world applications.
no code implementations • 22 Feb 2022 • Takuya Yashima, Takuya Narihira, Tamaki Kojima
In recent advances of deep generative models, face reenactment -manipulating and controlling human face, including their head movement-has drawn much attention for its wide range of applicability.
1 code implementation • 22 Mar 2021 • Kenji Suzuki, Yoshiyuki Kobayashi, Takuya Narihira
Moreover, our simple and general proposed method to calculate influence scores is available on our auto ML tool without programing, Neural Network Console.
no code implementations • 6 Mar 2021 • Andrew Shin, Masato Ishii, Takuya Narihira
Transformer architectures have brought about fundamental changes to computational linguistic field, which had been dominated by recurrent neural networks for many years.
1 code implementation • 12 Feb 2021 • Takuya Narihira, Javier Alonsogarcia, Fabien Cardinaux, Akio Hayakawa, Masato Ishii, Kazunori Iwaki, Thomas Kemp, Yoshiyuki Kobayashi, Lukas Mauch, Akira Nakamura, Yukio Obuchi, Andrew Shin, Kenji Suzuki, Stephen Tiedmann, Stefan Uhlich, Takuya Yashima, Kazuki Yoshiyama
While there exist a plethora of deep learning tools and frameworks, the fast-growing complexity of the field brings new demands and challenges, such as more flexible network design, speedy computation on distributed setting, and compatibility between different tools.
no code implementations • 25 Nov 2020 • Naofumi Akimoto, Akio Hayakawa, Andrew Shin, Takuya Narihira
To address this issue, we warp colors only from the regions on the reference frame restricted by correspondence in time.
no code implementations • 27 Oct 2020 • Akio Hayakawa, Takuya Narihira
We propose a novel out-of-core algorithm that enables faster training of extremely large-scale neural networks with sizes larger than allotted GPU memory.
no code implementations • 9 Aug 2019 • Yuka Ariki, Takuya Narihira
Path-planning algorithms are an important part of a wide variety of robotic applications, such as mobile robot navigation and robot arm manipulation.
no code implementations • CVPR 2016 • Michael Maire, Takuya Narihira, Stella X. Yu
Spectral embedding provides a framework for solving perceptual organization problems, including image segmentation and figure/ground organization.
no code implementations • ICCV 2015 • Takuya Narihira, Michael Maire, Stella X. Yu
We demonstrate results on both the synthetic images of Sintel and the real images of the classic MIT intrinsic image dataset.
no code implementations • 21 Nov 2015 • Takuya Narihira, Damian Borth, Stella X. Yu, Karl Ni, Trevor Darrell
We consider the visual sentiment task of mapping an image to an adjective noun pair (ANP) such as "cute baby".
no code implementations • CVPR 2015 • Takuya Narihira, Michael Maire, Stella X. Yu
We develop a new approach to inferring lightness, the perceived reflectance of surfaces, from a single image.