no code implementations • 29 Mar 2024 • Yuiko Sakuma, Masakazu Yoshimura, Junji Otsuka, Atsushi Irie, Takeshi Ohashi
To tackle these challenges, first, we study the effective search space design for fine-tuning a VFM by comparing different operators (such as resolution, feature size, width, depth, and bit-widths) in terms of performance and BitOPs reduction.
no code implementations • 15 Mar 2024 • Masakazu Yoshimura, Junji Otsuka, Takeshi Ohashi
Full DNN-based image signal processors (ISPs) have been actively studied and have achieved superior image quality compared to conventional ISPs.
Ranked #2 on Image Enhancement on MIT-Adobe 5k (PSNR on proRGB metric)
no code implementations • 24 Mar 2023 • Junji Otsuka, Masakazu Yoshimura, Takeshi Ohashi
To tackle these limitations, we propose a self-supervised reversed ISP method that does not require metadata and paired images.
no code implementations • 9 Nov 2022 • Siddharth Sagar Nijhawan, Leo Hoshikawa, Atsushi Irie, Masakazu Yoshimura, Junji Otsuka, Takeshi Ohashi
We propose a light-weight and highly efficient Joint Detection and Tracking pipeline for the task of Multi-Object Tracking using a fully-transformer architecture.
no code implementations • ICCV 2023 • Masakazu Yoshimura, Junji Otsuka, Atsushi Irie, Takeshi Ohashi
Image Signal Processors (ISPs) play important roles in image recognition tasks as well as in the perceptual quality of captured images.
no code implementations • CVPR 2023 • Masakazu Yoshimura, Junji Otsuka, Atsushi Irie, Takeshi Ohashi
We show that our proposed noise-accounted RAW augmentation method doubles the image recognition accuracy in challenging environments only with simple training data.
no code implementations • 15 Mar 2021 • Masakazu Yoshimura, Murilo Marques Marinho, Kanako Harada, Mamoru Mitsuishi
Our experiments demonstrated an improvement of 21 % for translation error and 26 % for orientation error on synthetic test data with respect to our previous work.
no code implementations • 15 Oct 2020 • Masakazu Yoshimura, Satoshi Ogata
Usually, such situations were solved by two separate models; one is a face detector model which crops facial regions and the other is an age estimation model which estimates from cropped images.
no code implementations • 3 Mar 2020 • Masakazu Yoshimura, Murilo M. Marinho, Kanako Harada, Mamoru Mitsuishi
To avoid injuring the patients, a collision-avoidance algorithm that depends on having an accurate model for the poses of the instruments' shafts is used.