no code implementations • 13 Sep 2022 • Vy Nguyen, Masayuki Tanaka, Yusuke Monno, Masatoshi Okutomi
Polarization information of light in a scene is valuable for various image processing and computer vision tasks.
1 code implementation • 5 Jul 2022 • Ikuro Sato, Ryota Yamada, Masayuki Tanaka, Nakamasa Inoue, Rei Kawakami
We developed a training algorithm called PoF: Post-Training of Feature Extractor that updates the feature extractor part of an already-trained deep model to search a flatter minimum.
no code implementations • 23 Jul 2021 • Napat Wanchaitanawong, Masayuki Tanaka, Takashi Shibata, Masatoshi Okutomi
Due to this assumption's breakdown, the position of the bounding boxes does not match between the two modalities, resulting in a significant decrease in detection accuracy, especially in regions where the amount of misalignment is large.
no code implementations • 22 Jul 2021 • Takashi Shibata, Masayuki Tanaka, Masatoshi Okutomi
The performance of CNN is dramatically degraded by geometric transformation, such as large rotations.
no code implementations • 14 Jan 2021 • Masayuki Tanaka, Hiroyuki Ikeda, Kazumi Murata, Satoshi Takita, Sogo Mineo, Michitaro Koike, Yuki Okura, Sumiko Harasawa
HSC is an optical wide-field imager installed at the prime focus of the Subaru Telescope and has been in operation since 2014.
Instrumentation and Methods for Astrophysics Astrophysics of Galaxies
no code implementations • 20 Nov 2020 • Takeru Suda, Masayuki Tanaka, Yusuke Monno, Masatoshi Okutomi
In this paper, we propose a deep snapshot high dynamic range (HDR) imaging framework that can effectively reconstruct an HDR image from the RAW data captured using a multi-exposure color filter array (ME-CFA), which consists of a mosaic pattern of RGB filters with different exposure levels.
no code implementations • 13 Nov 2020 • Wonjik Kim, Masayuki Tanaka, Masatoshi Okutomi, Yoko SASAKI
A common limitation of the existing learning-based approaches is a mismatch of training data and test data.
1 code implementation • 16 Oct 2020 • Tao Zhong, Wonjik Kim, Masayuki Tanaka, Masatoshi Okutomi
It has a two-branch structure, i. e., the spatial segmentation branch and the temporal velocity estimation branch.
no code implementations • 28 Jul 2020 • Miki Morimatsu, Yusuke Monno, Masayuki Tanaka, Masatoshi Okutomi
Since the polarimeter consists of an image sensor equipped with a monochrome or color polarization filter array (MPFA or CPFA), the demosaicking process to interpolate missing pixel values plays a crucial role in obtaining high-quality polarization images.
2 code implementations • 20 Jul 2020 • Wonjik Kim, Asako Kanezaki, Masayuki Tanaka
The usage of convolutional neural networks (CNNs) for unsupervised image segmentation was investigated in this study.
no code implementations • 15 Jun 2020 • Kazuki Endo, Masayuki Tanaka, Masatoshi Okutomi
Classification for degraded images having various levels of degradation is very important in practical applications.
no code implementations • 11 Mar 2020 • Wonjik Kim, Masayuki Tanaka, Masatoshi Okutomi, Yoko SASAKI
In this paper, we propose an automatic labeled sequential data generation pipeline for human segmentation and velocity estimation with point clouds.
no code implementations • 21 Jan 2020 • Koki Madono, Masayuki Tanaka, Masaki Onishi, Tetsuji Ogawa
In this study, a perceptually hidden object-recognition method is investigated to generate secure images recognizable by humans but not machines.
no code implementations • 12 Sep 2019 • Masanari Kimura, Masayuki Tanaka
Improving the interpretability of DNNs is one of the hot research topics.
no code implementations • 4 Jun 2019 • Ikuro Sato, Kohta Ishikawa, Guoqing Liu, Masayuki Tanaka
This study addresses an issue of co-adaptation between a feature extractor and a classifier in a neural network.
no code implementations • 7 May 2019 • Masanari Kimura, Masayuki Tanaka
To tackle this problem, we exploit a multi-channel attention mechanism in feature space.
no code implementations • 30 Apr 2019 • Masanari Kimura, Masayuki Tanaka
To tackle this problem, we exploit a multi-channel attention mechanism in feature space.
no code implementations • 13 Mar 2019 • Hiroshi Kuwajima, Masayuki Tanaka, Masatoshi Okutomi
However, the inference process of deep learning is black-box, and not very suitable to safety-critical systems which must exhibit high transparency.
1 code implementation • 14 Feb 2019 • Wonjik Kim, Masayuki Tanaka, Masatoshi Okutomi, Yoko SASAKI
We present 500k+ data generated by the proposed pipeline.
no code implementations • 23 Dec 2018 • Kazutaka Uchida, Masayuki Tanaka, Masatoshi Okutomi
However, image degradations in practice are often a mixture of several types of degradation.
no code implementations • 3 Oct 2018 • Masayuki Tanaka
In this paper, a weighted sigmoid gate unit (WiG) is proposed as the activation function.
no code implementations • 25 Sep 2018 • Masayuki Tanaka, Takashi Shibata, Masatoshi Okutomi
A low-light image enhancement is a highly demanded image processing technique, especially for consumer digital cameras and cameras on mobile phones.
no code implementations • 11 Sep 2018 • Kazutaka Uchida, Masayuki Tanaka, Masatoshi Okutomi
Blind image restoration processors based on convolutional neural network (CNN) are intensively researched because of their high performance.
no code implementations • 19 Mar 2018 • Masayuki Tanaka
This scheme allows us to train the network without the privacy issues.
no code implementations • 7 Dec 2017 • Hiroshi Kuwajima, Masayuki Tanaka
Safety critical systems strongly require the quality aspects of artificial intelligence including explainability.
no code implementations • ICCV 2017 • Takashi Shibata, Masayuki Tanaka, Masatoshi Okutomi
Next, the joint-filter cost volume and a set of filtered images are computed from the target image and the set of the translated guidances.
no code implementations • 20 Apr 2017 • Masayuki Tanaka, Jean Coupon, Bau-Ching Hsieh, Sogo Mineo, Atsushi J. Nishizawa, Joshua Speagle, Hisanori Furusawa, Satoshi Miyazaki, Hitoshi Murayama
Photometric redshifts are a key component of many science objectives in the Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP).
Astrophysics of Galaxies
no code implementations • CVPR 2016 • Takashi Shibata, Masayuki Tanaka, Masatoshi Okutomi
To generate detail-preserving and artifact-free output images, we combine the benefits of the two approaches into the proposed framework by introducing the intensity-range constraint and the base-structure constraint.