no code implementations • 15 Apr 2024 • Tianhan Xu, Takuya Ikeda, Koichi Nishiwaki
In this paper, we propose a method to segment and recover a static, clean background and multiple 360$^\circ$ objects from observations of scenes at different timestamps.
no code implementations • 20 Feb 2024 • Takuya Ikeda, Sergey Zakharov, Tianyi Ko, Muhammad Zubair Irshad, Robert Lee, Katherine Liu, Rares Ambrus, Koichi Nishiwaki
This paper addresses the challenging problem of category-level pose estimation.
no code implementations • 23 Nov 2023 • Pengyuan Wang, Takuya Ikeda, Robert Lee, Koichi Nishiwaki
This requires significantly less data to train than prior methods since the semantic features are robust to object texture and appearance.
no code implementations • 30 May 2023 • Hiroya Sato, Takuya Ikeda, Koichi Nishiwaki
In order to handle the ambiguity, the Bingham distribution is one promising solution.
no code implementations • 9 Mar 2022 • Hiroya Sato, Takuya Ikeda, Koichi Nishiwaki
In recent years, a deep learning framework has been widely used for object pose estimation.
no code implementations • 3 Mar 2022 • Takuya Ikeda, Suomi Tanishige, Ayako Amma, Michael Sudano, Hervé Audren, Koichi Nishiwaki
This improves the quality of synthetic data for training pose estimation networks.