ColANet: A UAV Collision Avoidance Dataset

Pedro, Dário; Mora, André; Carvalho, João Paulo; Azevedo, Fábio Mícolis de; Fonseca, José

doi:10.1007/978-3-030-45124-0_5

Cited by 16 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To train the proposed NNP, the creation of a new dataset focused on the selected testing scenario, the avoidance of a thrown ball, was required. The techniques proposed in ColANet [57] were used as guidelines, and the new dataset was made available at https://ballnet.qa.pdmfc.com/ (accessed on 20 May 2021). This dataset has 600 videos, which represent a total of 20,000 images.…”

Section: Nnp Training and Resultsmentioning

confidence: 99%

Collision Avoidance on Unmanned Aerial Vehicles Using Neural Network Pipelines and Flow Clustering Techniques

et al. 2021

Self Cite

View full text Add to dashboard Cite

Unmanned Autonomous Vehicles (UAV), while not a recent invention, have recently acquired a prominent position in many industries, and they are increasingly used not only by avid customers, but also in high-demand technical use-cases, and will have a significant societal effect in the coming years. However, the use of UAVs is fraught with significant safety threats, such as collisions with dynamic obstacles (other UAVs, birds, or randomly thrown objects). This research focuses on a safety problem that is often overlooked due to a lack of technology and solutions to address it: collisions with non-stationary objects. A novel approach is described that employs deep learning techniques to solve the computationally intensive problem of real-time collision avoidance with dynamic objects using off-the-shelf commercial vision sensors. The suggested approach’s viability was corroborated by multiple experiments, firstly in simulation, and afterward in a concrete real-world case, that consists of dodging a thrown ball. A novel video dataset was created and made available for this purpose, and transfer learning was also tested, with positive results.

show abstract

Section: Nnp Training and Resultsmentioning

confidence: 99%

Collision Avoidance on Unmanned Aerial Vehicles Using Neural Network Pipelines and Flow Clustering Techniques

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…To train the proposed architecture, the dataset ColANet (at https://colanet.qa.pdmfc.com/) was used [22], which contains around 100 videos that result to 18,872 images. This is a video dataset of collisions and has the possibility to output a classification target (collision or no collision) or a regression target.…”

Section: Training and Resultsmentioning

confidence: 99%

“…Furthermore, a novel Dynamic Collision Avoidance algorithm is proposed, which utilizes a Convolutional Neural Network (CNN) to extract features from video frames, and a Recursive Neural Network (RNN) that takes advantage of the video temporal characteristics capable of estimating if there is an incoming collision, such as the one depicted in Figure 1. This figure illustrates two frames from a video, retrieved from the ColANet dataset [22], that presents a kid kicking a ball into a UAV. There are quite a few attempts to integrate UAVs into an architecture that links different UAVs modules and a cloud platform.…”

Section: Introductionmentioning

confidence: 99%

FFAU—Framework for Fully Autonomous UAVs

et al. 2020

Self Cite

View full text Add to dashboard Cite

Unmanned Aerial Vehicles (UAVs), although hardly a new technology, have recently gained a prominent role in many industries being widely used not only among enthusiastic consumers, but also in high demanding professional situations, and will have a massive societal impact over the coming years. However, the operation of UAVs is fraught with serious safety risks, such as collisions with dynamic obstacles (birds, other UAVs, or randomly thrown objects). These collision scenarios are complex to analyze in real-time, sometimes being computationally impossible to solve with existing State of the Art (SoA) algorithms, making the use of UAVs an operational hazard and therefore significantly reducing their commercial applicability in urban environments. In this work, a conceptual framework for both stand-alone and swarm (networked) UAVs is introduced, with a focus on the architectural requirements of the collision avoidance subsystem to achieve acceptable levels of safety and reliability. The SoA principles for collision avoidance against stationary objects are reviewed and a novel approach is described, using deep learning techniques to solve the computational intensive problem of real-time collision avoidance with dynamic objects. The proposed framework includes a web-interface allowing the full control of UAVs as remote clients with a supervisor cloud-based platform. The feasibility of the proposed approach was demonstrated through experimental tests using a UAV, developed from scratch using the proposed framework. Test flight results are presented for an autonomous UAV monitored from multiple countries across the world.

show abstract

“…The dataset consists of 10k RGB images used to estimate the depth and therefore allow safe navigation of a ground robot. More recently, [15] released a collision avoidance video dataset collected by means of a drone equipped with a frontward RGB camera. A distinctive feature of this dataset is that it includes 100 samples of actual collisions with static and moving obstacles.…”

Section: Introductionmentioning

confidence: 99%

A Novel Obstacle Detection and Avoidance Dataset for Drones

Dupeyroux

Dinaux

Wessendorp³

et al. 2022

System Engineering for Constrained Embedded Systems

View full text Add to dashboard Cite

In this paper, we introduce the Obstacle Detection & Avoidance (ODA) Dataset for Drones, aiming at providing raw data obtained in a real indoor environment with sensors adapted for aerial robotics in the context of obstacle detection and avoidance. Our micro air vehicle (MAV) is equipped with the following sensors: (i) an event-based camera, the performance of which makes it optimized for drone applications; (ii) a standard RGB camera; (iii) a 24-GHz radar sensor to enhance multi-sensory solutions; and (iv) a 6-axes IMU. The ground truth position and attitude are provided by an OptiTrack motion capture system. The resulting dataset consists of more than 1350 sequences obtained in four distinct conditions (one or two obstacles, full or dim light). It is intended for benchmarking algorithmic and neural solutions for obstacle detection and avoidance with UAVs, but also course estimation and in general autonomous navigation. The dataset is available at: https://github.com/tudelft/ODA_Dataset [6]. CCS CONCEPTS• Computer systems organization → Robotic autonomy; Robotic control; Sensors and actuators.

show abstract

ColANet: A UAV Collision Avoidance Dataset

Cited by 16 publications

References 29 publications

Collision Avoidance on Unmanned Aerial Vehicles Using Neural Network Pipelines and Flow Clustering Techniques

Collision Avoidance on Unmanned Aerial Vehicles Using Neural Network Pipelines and Flow Clustering Techniques

FFAU—Framework for Fully Autonomous UAVs

A Novel Obstacle Detection and Avoidance Dataset for Drones

Contact Info

Product

Resources

About