scispace - formally typeset
Proceedings ArticleDOI

Learning to fly by crashing

Reads0
Chats0
TLDR
This paper builds a drone whose sole purpose is to crash into objects: it samples naive trajectories and crashes into random objects to create one of the biggest UAV crash dataset.
Abstract
How do you learn to navigate an Unmanned Aerial Vehicle (UAV) and avoid obstacles? One approach is to use a small dataset collected by human experts: however, high capacity learning algorithms tend to overfit when trained with little data. An alternative is to use simulation. But the gap between simulation and real world remains large especially for perception problems. The reason most research avoids using large-scale real data is the fear of crashes! In this paper, we propose to bite the bullet and collect a dataset of crashes itself! We build a drone whose sole purpose is to crash into objects: it samples naive trajectories and crashes into random objects. We crash our drone 11,500 times to create one of the biggest UAV crash dataset. This dataset captures the different ways in which a UAV can crash. We use all this negative flying data in conjunction with positive data sampled from the same trajectories to learn a simple yet powerful policy for UAV navigation. We show that this simple self-supervised model is quite effective in navigating the UAV even in extremely cluttered environments with dynamic obstacles including humans. For supplementary video see:

read more

Citations
More filters
Book ChapterDOI

Spherical Fully Covered UAV with Autonomous Indoor Localization

TL;DR: A UAV (Unmanned Aerial Vehicle) with intrinsic safety which can interact with people and obstacles while flying in an indoor environment in an autonomous way is presented.
Posted Content

Fully Onboard AI-powered Human-Drone Pose Estimation on Ultra-low Power Autonomous Flying Nano-UAVs

TL;DR: In this article, a parallel ultra-low-power (PULP) architecture is used for the real-time execution of a deep neural network (DNN) on a nano-UAV.
DissertationDOI

Time-ordered Online Training of Convolutional Neural Networks

Ervin Teng
TL;DR: A metric, incremental training benefit (ITB) per annotation, is defined, that seeks to capture the value extracted for each annotation provided by the user, that is able to increase average ITB by 10-14 times versus a random search.
Journal ArticleDOI

Robotic Elytra: Insect-Inspired Protective Wings for Resilient and Multi-Modal Drones

TL;DR: In this paper, an additional set of retractable wings, named elytra, is added to the main folding wings when protection is needed, which can rapidly encapsulate the main foldable wings when needed.
Proceedings ArticleDOI

Learning visual policies for building 3D shape categories

TL;DR: In this paper, a disassembly procedure is proposed to discover new object instances and their assembly plans in state space and then rendered simulated states in the observation space and learn a heatmap representation to predict alternative actions from a given input image.
References
More filters
Proceedings Article

ImageNet Classification with Deep Convolutional Neural Networks

TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Journal ArticleDOI

Human-level control through deep reinforcement learning

TL;DR: This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
Proceedings ArticleDOI

Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation

TL;DR: RCNN as discussed by the authors combines CNNs with bottom-up region proposals to localize and segment objects, and when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost.
Posted Content

Rich feature hierarchies for accurate object detection and semantic segmentation

TL;DR: This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.
Proceedings ArticleDOI

Parallel Tracking and Mapping for Small AR Workspaces

TL;DR: A system specifically designed to track a hand-held camera in a small AR workspace, processed in parallel threads on a dual-core computer, that produces detailed maps with thousands of landmarks which can be tracked at frame-rate with accuracy and robustness rivalling that of state-of-the-art model-based systems.
Related Papers (5)