paper-with-me

Papers

UAS Navigation in the Real World Using Visual Observation

2022-08-25 · Yuci Han, Jianli Wei, Alper Yilmaz

This paper presents a novel end-to-end Unmanned Aerial System (UAS) navigation approach for long-range visual navigation in the real world. Inspired by dual-process visual navigation system of human's instinct: environment understanding and landmark recognition, we formulate the UAS navigation task into two same phases. Our system combines the reinforcement learning (RL) and image matching approaches. First, the agent learns the navigation policy using RL in the specified environment. To achieve this, we design an interactive UASNAV environment for the training process. Once the agent learns the navigation policy, which means 'familiarized themselves with the environment', we let the UAS fly in the real world to recognize the landmarks using image matching method and take action according to the learned policy. During the navigation process, the UAS is embedded with single camera as the only visual sensor. We demonstrate that the UAS can learn navigating to the destination hundreds meters away from the starting point with the shortest path in the real world scenario.

📄 PDF Abstract BibTeX arXiv:2208.12125

Code (0)

등록된 구현이 없습니다.

Tasks

Landmark RecognitionReinforcement Learning (RL)Visual Navigation

Similar Papers 제목 키워드 기반

On Embodied Visual Navigation in Real Environments Through Habitat

2020-10-26 · Marco Rosano, Antonino Furnari, Luigi Gulino, Giovanni Maria Farinella

Visual navigation models based on deep learning can learn effective policies when trained on large amounts of visual observations through reinforcement learning. Unfortunately, collecting the required experience in the r…

Domain AdaptationUnsupervised Domain AdaptationVisual Navigation

Ground-level Viewpoint Vision-and-Language Navigation in Continuous Environments

2025-02-26 · Zerui Li, Gengze Zhou, Haodong Hong, Yanyan Shao 외

Vision-and-Language Navigation (VLN) empowers agents to associate time-sequenced visual observations with corresponding instructions to make sequential decisions. However, generalization remains a persistent challenge, p…

Instruction FollowingVision and Language Navigation

NavWAM: A Navigation World Action Model for Goal-Conditioned Visual Navigation

2026-06-11 · Daichi Azuma, Taiki Miyanishi, Koya Sakamoto, Shuhei Kurita 외 arxiv

Goal-conditioned visual navigation requires a robot to act under partial observability by anticipating how its motion will change the future egocentric view and whether that change brings it closer to the goal. Navigatio…

Visual Navigation

Pathdreamer: A World Model for Indoor Navigation

2021-05-18 · ICCV 2021 10 · Jing Yu Koh, Honglak Lee, Yinfei Yang, Jason Baldridge 외

People navigating in unfamiliar buildings take advantage of myriad visual, spatial and semantic cues to efficiently achieve their navigation goals. Towards equipping computational agents with similar capabilities, we int…

modelSemantic SegmentationVision and Language Navigation

VELMA: Verbalization Embodiment of LLM Agents for Vision and Language Navigation in Street View

2023-07-12 · Raphael Schumann, Wanrong Zhu, Weixi Feng, Tsu-Jui Fu 외

Incremental decision making in real-world environments is one of the most challenging tasks in embodied artificial intelligence. One particularly demanding scenario is Vision and Language Navigation~(VLN) which requires …

Decision MakingNatural Language UnderstandingVision and Language Navigation