Learning to fly by crashing
01 Sep 2017-pp 3948-3955
TL;DR: 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:
Citations
More filters
TL;DR: The general principles of DL and convolutional neural networks are introduced, five major areas of application of DL in medical imaging and radiation therapy are surveyed, common themes are identified, methods for dataset expansion are discussed, and lessons learned, remaining challenges, and future directions are summarized.
Abstract: The goals of this review paper on deep learning (DL) in medical imaging and radiation therapy are to (a) summarize what has been achieved to date; (b) identify common and unique challenges, and strategies that researchers have taken to address these challenges; and (c) identify some of the promising avenues for the future both in terms of applications as well as technical innovations. We introduce the general principles of DL and convolutional neural networks, survey five major areas of application of DL in medical imaging and radiation therapy, identify common themes, discuss methods for dataset expansion, and conclude by summarizing lessons learned, remaining challenges, and future directions.
525 citations
23 Jan 2018
TL;DR: The proposed DroNet is a convolutional neural network that can safely drive a drone through the streets of a city, trained from data collected by cars and bicycles, which, already integrated into the urban environment, would not endanger other vehicles and pedestrians.
Abstract: Civilian drones are soon expected to be used in a wide variety of tasks, such as aerial surveillance, delivery, or monitoring of existing architectures. Nevertheless, their deployment in urban environments has so far been limited. Indeed, in unstructured and highly dynamic scenarios, drones face numerous challenges to navigate autonomously in a feasible and safe way. In contrast to traditional “map-localize-plan” methods, this letter explores a data-driven approach to cope with the above challenges. To accomplish this, we propose DroNet: a convolutional neural network that can safely drive a drone through the streets of a city. Designed as a fast eight-layers residual network, DroNet produces two outputs for each single input image: A steering angle to keep the drone navigating while avoiding obstacles, and a collision probability to let the UAV recognize dangerous situations and promptly react to them. The challenge is however to collect enough data in an unstructured outdoor environment such as a city. Clearly, having an expert pilot providing training trajectories is not an option given the large amount of data required and, above all, the risk that it involves for other vehicles or pedestrians moving in the streets. Therefore, we propose to train a UAV from data collected by cars and bicycles, which, already integrated into the urban environment, would not endanger other vehicles and pedestrians. Although trained on city streets from the viewpoint of urban vehicles, the navigation policy learned by DroNet is highly generalizable. Indeed, it allows a UAV to successfully fly at relative high altitudes and even in indoor environments, such as parking lots and corridors. To share our findings with the robotics community, we publicly release all our datasets, code, and trained networks.
382 citations
Cites background or methods or result from "Learning to fly by crashing"
...For this approach, we implemented the algorithm proposed in [10], which was shown to have very good control capabilities in indoor environments....
[...]
...effective flying policies [6], [9], [10], but they still leave the issue of collecting enough expert trajectories to imitate....
[...]
...Recall that [10] uses only collision probabilities, while DroNet uses also predicted steering angles, too....
[...]
...tionally, as pointed out by [10], collision trajectories avoided by expert human pilots are actually necessary to let the robotic platform learn how to behave in dangerous situations....
[...]
...jectories [10], or model predictive control [12]....
[...]
20 May 2019
TL;DR: In this paper, the authors demonstrate the first application of deep reinforcement learning to autonomous driving using a single monocular image as input, and provide a general and easy to obtain reward: the distance travelled by the vehicle without the safety driver taking control.
Abstract: We demonstrate the first application of deep reinforcement learning to autonomous driving. From randomly initialised parameters, our model is able to learn a policy for lane following in a handful of training episodes using a single monocular image as input. We provide a general and easy to obtain reward: the distance travelled by the vehicle without the safety driver taking control. We use a continuous, model-free deep reinforcement learning algorithm, with all exploration and optimisation performed on-vehicle. This demonstrates a new framework for autonomous driving which moves away from reliance on defined logical rules, mapping, and direct supervision. We discuss the challenges and opportunities to scale this approach to a broader range of autonomous driving tasks.
318 citations
Posted Content•
TL;DR: This work demonstrates a new framework for autonomous driving which moves away from reliance on defined logical rules, mapping, and direct supervision and provides a general and easy to obtain reward: the distance travelled by the vehicle without the safety driver taking control.
Abstract: We demonstrate the first application of deep reinforcement learning to autonomous driving. From randomly initialised parameters, our model is able to learn a policy for lane following in a handful of training episodes using a single monocular image as input. We provide a general and easy to obtain reward: the distance travelled by the vehicle without the safety driver taking control. We use a continuous, model-free deep reinforcement learning algorithm, with all exploration and optimisation performed on-vehicle. This demonstrates a new framework for autonomous driving which moves away from reliance on defined logical rules, mapping, and direct supervision. We discuss the challenges and opportunities to scale this approach to a broader range of autonomous driving tasks.
291 citations
Cites background from "Learning to fly by crashing"
...Variants of this approach have been demonstrated on self-driving cars [18], [19] and drones [20], [21], [22]....
[...]
Posted Content•
TL;DR: It is demonstrated that visual MPC can generalize to never-before-seen objects---both rigid and deformable---and solve a range of user-defined object manipulation tasks using the same model.
Abstract: Deep reinforcement learning (RL) algorithms can learn complex robotic skills from raw sensory inputs, but have yet to achieve the kind of broad generalization and applicability demonstrated by deep learning methods in supervised domains. We present a deep RL method that is practical for real-world robotics tasks, such as robotic manipulation, and generalizes effectively to never-before-seen tasks and objects. In these settings, ground truth reward signals are typically unavailable, and we therefore propose a self-supervised model-based approach, where a predictive model learns to directly predict the future from raw sensory readings, such as camera images. At test time, we explore three distinct goal specification methods: designated pixels, where a user specifies desired object manipulation tasks by selecting particular pixels in an image and corresponding goal positions, goal images, where the desired goal state is specified with an image, and image classifiers, which define spaces of goal states. Our deep predictive models are trained using data collected autonomously and continuously by a robot interacting with hundreds of objects, without human supervision. We demonstrate that visual MPC can generalize to never-before-seen objects---both rigid and deformable---and solve a range of user-defined object manipulation tasks using the same model.
281 citations
Cites background from "Learning to fly by crashing"
...A number of recent works have studied self-supervised robotic learning, where large-scale unattended data collection is used to learn individual skills such as grasping [22], [23], [24], [25], pushgrasp synergies [26], or obstacle avoidance [27], [28]....
[...]
References
More filters
Proceedings Article•
03 Dec 2012TL;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.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, 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. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
73,978 citations
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.
Abstract: The theory of reinforcement learning provides a normative account, deeply rooted in psychological and neuroscientific perspectives on animal behaviour, of how agents may optimize their control of an environment. To use reinforcement learning successfully in situations approaching real-world complexity, however, agents are confronted with a difficult task: they must derive efficient representations of the environment from high-dimensional sensory inputs, and use these to generalize past experience to new situations. Remarkably, humans and other animals seem to solve this problem through a harmonious combination of reinforcement learning and hierarchical sensory processing systems, the former evidenced by a wealth of neural data revealing notable parallels between the phasic signals emitted by dopaminergic neurons and temporal difference reinforcement learning algorithms. While reinforcement learning agents have achieved some successes in a variety of domains, their applicability has previously been limited to domains in which useful features can be handcrafted, or to domains with fully observed, low-dimensional state spaces. Here we use recent advances in training deep neural networks to develop a novel artificial agent, termed a deep Q-network, that can learn successful policies directly from high-dimensional sensory inputs using end-to-end reinforcement learning. We tested this agent on the challenging domain of classic Atari 2600 games. We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to surpass the performance of all previous algorithms and achieve a level comparable to that of a professional human games tester across a set of 49 games, using the same algorithm, network architecture and hyperparameters. 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.
23,074 citations
"Learning to fly by crashing" refers background in this paper
...Deep reinforcement learning methods [20] have shown impressive results, however they are too data intensive (order of million examples) for our task of drone flying....
[...]
23 Jun 2014
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.
Abstract: Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine multiple low-level image features with high-level context. In this paper, we propose 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%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also present experiments that provide insight into what the network learns, revealing a rich hierarchy of image features. Source code for the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.
21,729 citations
Posted Content•
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%.
Abstract: Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine multiple low-level image features with high-level context. In this paper, we propose 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%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also compare R-CNN to OverFeat, a recently proposed sliding-window detector based on a similar CNN architecture. We find that R-CNN outperforms OverFeat by a large margin on the 200-class ILSVRC2013 detection dataset. Source code for the complete system is available at this http URL.
13,081 citations
"Learning to fly by crashing" refers background or methods in this paper
...We use ImageNet-pretrained weights as initialization for our network [22]....
[...]
...A key component of deep learning, is the high amount of data required to train these generalizable models [21], [22]....
[...]
13 Nov 2007
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.
Abstract: This paper presents a method of estimating camera pose in an unknown scene. While this has previously been attempted by adapting SLAM algorithms developed for robotic exploration, we propose a system specifically designed to track a hand-held camera in a small AR workspace. We propose to split tracking and mapping into two separate tasks, processed in parallel threads on a dual-core computer: one thread deals with the task of robustly tracking erratic hand-held motion, while the other produces a 3D map of point features from previously observed video frames. This allows the use of computationally expensive batch optimisation techniques not usually associated with real-time operation: The result is a system that produces detailed maps with thousands of landmarks which can be tracked at frame-rate, with an accuracy and robustness rivalling that of state-of-the-art model-based systems.
4,091 citations
"Learning to fly by crashing" refers methods in this paper
...For this purpose, we use PTAM [27] module that localizes the robot and helps it backtrack....
[...]
...For this purpose, we use PTAM [26] module that localizes the robot and helps it backtrack....
[...]
...Algorithm 1 Data Collection 1: Init: Track position error using PTAM from intial Take Off location 2: while No Crash do 3: Choose a random direction 4: while No Collision do 5: Continue in the same direction 6: while Position error > do 7: calculate control command based on position error Algorithm 1 describes our method of data collection in more detail....
[...]