paper-with-me

홈 › Papers

Use of First and Third Person Views for Deep Intersection Classification

2019-01-22 · Koji Takeda, Kanji Tanaka

We explore the problem of intersection classification using monocular on-board passive vision, with the goal of classifying traffic scenes with respect to road topology. We divide the existing approaches into two broad categories according to the type of input data: (a) first person vision (FPV) approaches, which use an egocentric view sequence as the intersection is passed; and (b) third person vision (TPV) approaches, which use a single view immediately before entering the intersection. The FPV and TPV approaches each have advantages and disadvantages. Therefore, we aim to combine them into a unified deep learning framework. Experimental results show that the proposed FPV-TPV scheme outperforms previous methods and only requires minimal FPV/TPV measurements.

📄 PDF Abstract BibTeX arXiv:1901.07446

Code (0)

등록된 구현이 없습니다.

Tasks

General Classification

Similar Papers 제목 키워드 기반

Fusing Personal and Environmental Cues for Identification and Segmentation of First-Person Camera Wearers in Third-Person Views

2024-01-01 · CVPR 2024 1 · Ziwei Zhao, Yuchen Wang, Chuhua Wang

As wearable cameras become more popular an important question emerges: how to identify camera wearers within the perspective of conventional static cameras. The drastic difference between first-person (egocentric) an…

Exploring Self-Attention for Visual Intersection Classification

2022-03-26 · Haruki Nakata, Kanji Tanaka, Koji Takeda

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 captur…

Classification

Joint Person Segmentation and Identification in Synchronized First- and Third-person Videos

2018-03-29 · ECCV 2018 9 · Mingze Xu, Chenyou Fan, Yuchen Wang, Michael S. Ryoo 외

In a world of pervasive cameras, public spaces are often captured from multiple perspectives by cameras of different types, both fixed and mobile. An important problem is to organize these heterogeneous collections of vi…

Segmentation

Summarizing First-Person Videos from Third Persons' Points of Views

2017-11-24 · ECCV 2018 · Hsuan-I Ho, Wei-Chen Chiu, Yu-Chiang Frank Wang

Video highlight or summarization is among interesting topics in computer vision, which benefits a variety of applications like viewing, searching, or storage. However, most existing studies rely on training data of third…

POV: Prompt-Oriented View-Agnostic Learning for Egocentric Hand-Object Interaction in the Multi-View World

2024-03-09 · Boshen Xu, Sipeng Zheng, Qin Jin

We humans are good at translating third-person observations of hand-object interactions (HOI) into an egocentric view. However, current methods struggle to replicate this ability of view adaptation from third-person to f…