We present AVOD, an Aggregate View Object Detection network for autonomous driving scenarios. The proposed neural network architecture uses LIDAR point clouds and RGB images to generate features that are shared by two subnetworks: a region proposal network (RPN) and a second stage detector network. The proposed RPN uses a novel architecture capable of performing multimodal feature fusion on high resolution feature maps to generate reliable 3D object proposals for multiple object classes in road scenes. Using these proposals, the second stage detection network performs accurate oriented 3D bounding box regression and category classification to predict the extents, orientation, and classification of objects in 3D space. Our proposed architecture is shown to produce state of the art results on the KITTI 3D object detection benchmark [1] while running in real time with a low memory footprint, making it a suitable candidate for deployment on autonomous vehicles. Code is at: https://github.com/kujason/avod

No abstract

The Internet of Drones (IoD) is a layered network control architecture designed mainly for coordinating the access of unmanned aerial vehicles to controlled airspace, and providing navigation services between locations referred to as nodes. The IoD provides generic services for various drone applications such as package delivery, traffic surveillance, search and rescue and more. In this paper, we present a conceptual model of how such an architecture can be organized and we specify the features that an IoD system based on our architecture should implement. For doing so, we extract key concepts from three existing large scale networks, namely the air traffic control network, the cellular network, and the Internet and explore their connections to our novel architecture for drone traffic management. A simulation platform for IoD is being implemented which can be accessed from www.IoDnet.org in the future.Index Terms-Layered architecture, Internet of Drones (IoD), Internet, cellular network, air traffic control (ATC), low altitude air traffic management, unmanned aerial vehicle (UAV). trical engineering from Sharif University of Technology. He completed his M.Math, and currently is pursuing a PhD degree both in He is also the Director of the Waterloo Autonomous Vehicles Laboratory. His main research interests include perception, navigation and control of autonomous aerial rotorcraft and ground rovers with a focus on simultaneous localization and mapping, optimal motion planning and multi-robot coordination.

No abstract

No abstract

With the rise of data driven deep neural networks as a realization of universal function approximators, most research on computer vision problems has moved away from hand crafted classical image processing algorithms. This paper shows that with a well designed algorithm, we are capable of outperforming neural network based methods on the task of depth completion. The proposed algorithm is simple and fast, runs on the CPU, and relies only on basic image processing operations to perform depth completion of sparse LIDAR depth data. We evaluate our algorithm on the challenging KITTI depth completion benchmark [1], and at the time of submission, our method ranks f irst on the KITTI test server among all published methods. Furthermore, our algorithm is data independent, requiring no training data to perform the task at hand. The code written in Python will be made publicly available at https://github.com/kujason/ip basic.

We present MonoPSR, a monocular 3D object detection method that leverages proposals and shape reconstruction. First, using the fundamental relations of a pinhole camera model, detections from a mature 2D object detector are used to generate a 3D proposal per object in a scene. The 3D location of these proposals prove to be quite accurate, which greatly reduces the difficulty of regressing the final 3D bounding box detection. Simultaneously, a point cloud is predicted in an object centered coordinate system to learn local scale and shape information. However, the key challenge is how to exploit shape information to guide 3D localization. As such, we devise aggregate losses, including a novel projection alignment loss, to jointly optimize these tasks in the neural network to improve 3D localization accuracy. We validate our method on the KITTI benchmark where we set new state-of-the-art results among published monocular methods, including the harder pedestrian and cyclist classes, while maintaining efficient run-time.

No abstract