Výsledky vyhledávání - "TAKEDA, KOJI"

Report

Zero-shot Degree of Ill-posedness Estimation for Active Small Object Change Detection

Autor: Takeda, Koji, Tanaka, Kanji, Nakamura, Yoshimasa, Kanezaki, Asako

In everyday indoor navigation, robots often needto detect non-distinctive small-change objects (e.g., stationery,lost items, and junk, etc.) to maintain domain knowledge. Thisis most relevant to ground-view change detection (GVCD), a recently emergin

Externí odkaz: http://arxiv.org/abs/2405.06185

Zobrazit plný text záznamu

Report

Lifelong Change Detection: Continuous Domain Adaptation for Small Object Change Detection in Every Robot Navigation

Autor: Takeda, Koji, Tanaka, Kanji, Nakamura, Yoshimasa

The recently emerging research area in robotics, ground view change detection, suffers from its ill-posed-ness because of visual uncertainty combined with complex nonlinear perspective projection. To regularize the ill-posed-ness, the commonly applie

Externí odkaz: http://arxiv.org/abs/2306.16086

Zobrazit plný text záznamu

Report

Domain Invariant Siamese Attention Mask for Small Object Change Detection via Everyday Indoor Robot Navigation

Autor: Takeda, Koji, Tanaka, Kanji, Nakamura, Yoshimasa

The problem of image change detection via everyday indoor robot navigation is explored from a novel perspective of the self-attention technique. Detecting semantically non-distinctive and visually small changes remains a key challenge in the robotics

Externí odkaz: http://arxiv.org/abs/2203.15362

Zobrazit plný text záznamu

Report

Exploring Self-Attention for Visual Intersection Classification

Autor: Nakata, Haruki, Tanaka, Kanji, Takeda, Koji

In robot vision, self-attention has recently emerged as a technique for capturing non-local contexts. In this study, we introduced a self-attention mechanism into the intersection recognition system as a method to capture the non-local contexts behin

Externí odkaz: http://arxiv.org/abs/2203.13977

Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání