English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper presents our method for enabling a UAV quadrotor, equipped with amonocular camera, to autonomously avoid collisions with obstacles inunstructured and unknown indoor environments. When compared to obstacleavoidance in ground vehicular robots, UAV navigation brings in additionalchallenges because the UAV motion is no more constrained to a well-definedindoor ground or street environment. Horizontal structures in indoor andoutdoor environments like decorative items, furnishings, ceiling fans,sign-boards, tree branches etc., also become relevant obstacles unlike thosefor ground vehicular robots. Thus, methods of obstacle avoidance developed forground robots are clearly inadequate for UAV navigation. Current controlmethods using monocular images for UAV obstacle avoidance are heavily dependenton environment information. These controllers do not fully retain and utilizethe extensively available information about the ambient environment fordecision making. We propose a deep reinforcement learning based method for UAVobstacle avoidance (OA) and autonomous exploration which is capable of doingexactly the same. The crucial idea in our method is the concept of partialobservability and how UAVs can retain relevant information about theenvironment structure to make better future navigation decisions. Our OAtechnique uses recurrent neural networks with temporal attention and providesbetter results compared to prior works in terms of distance covered duringnavigation without collisions. In addition, our technique has a high inferencerate (a key factor in robotic applications) and is energy-efficient as itminimizes oscillatory motion of UAV and reduces power wastage.

Memory-based Deep Reinforcement Learning for Obstacle Avoidance in UAV with Limited Environment Knowledge