scholarly journals Reconfigurable Embedded Devices Using Reinforcement Learning to Develop Action Policies

2020 ◽  
Vol 15 (4) ◽  
pp. 1-25
Author(s):  
Alwyn Burger ◽  
Gregor Schiele ◽  
David W. King
Keyword(s):  
Neural Networks ◽  
Smart Cities ◽  
Waste Water Treatment ◽  
Work Load ◽  
Component Composition ◽  
Network Failure ◽  
Q Learning ◽  
Field Programmable ◽  
Reward Functions ◽  
Guide Device

The size of sensor networks supporting smart cities is ever increasing. Sensor network resiliency becomes vital for critical networks such as emergency response and waste water treatment. One approach is to engineer “self-aware” sensors that can proactively change their component composition in response to changes in work load when critical devices fail. By extension, these devices could anticipate their own termination, such as battery depletion, and offload current tasks onto connected devices. These neighboring devices can then reconfigure themselves to process these tasks, thus avoiding catastrophic network failure. In this article, we compare and contrast two types of self-aware sensors. One set uses Q-learning to develop a policy that guides device reaction to various environmental stimuli, whereas the others use a set of shallow neural networks to select an appropriate reaction. The novelty lies in the use of field programmable gate arrays embedded on the sensors that take into account internal system state, configuration, and learned state-action pairs, which guide device decisions to meet system demands. Experiments show that even relatively simple reward functions develop both Q-learning policies and shallow neural networks that yield positive device behaviors in dynamic environments.

scholarly journals Reinforcement Learning for Hyperparameter Tuning in Deep Learning-based Side-channel Analysis

2021 ◽  
pp. 677-707
Author(s):  
Jorai Rijsdijk ◽  
Lichao Wu ◽  
Guilherme Perin ◽  
Stjepan Picek
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Reinforcement Learning ◽  
Convolutional Neural Networks ◽  
Random Search ◽  
High Price ◽  
Side Channel ◽  
Q Learning ◽  
Reward Functions

Deep learning represents a powerful set of techniques for profiling sidechannel analysis. The results in the last few years show that neural network architectures like multilayer perceptron and convolutional neural networks give strong attack performance where it is possible to break targets protected with various countermeasures. Considering that deep learning techniques commonly have a plethora of hyperparameters to tune, it is clear that such top attack results can come with a high price in preparing the attack. This is especially problematic as the side-channel community commonly uses random search or grid search techniques to look for the best hyperparameters.In this paper, we propose to use reinforcement learning to tune the convolutional neural network hyperparameters. In our framework, we investigate the Q-Learning paradigm and develop two reward functions that use side-channel metrics. We mount an investigation on three commonly used datasets and two leakage models where the results show that reinforcement learning can find convolutional neural networks exhibiting top performance while having small numbers of trainable parameters. We note that our approach is automated and can be easily adapted to different datasets. Several of our newly developed architectures outperform the current state-of-the-art results. Finally, we make our source code publicly available. https://github.com/AISyLab/Reinforcement-Learning-for-SCA

scholarly journals LOW-COST FIELD PROGRAMMABLE GATE ARRAY ACCELERATES DEEP Q-LEARNING

2021 ◽  
Author(s):  
JINGHUI WANG ◽  
YUANCHAO ZHAO
Keyword(s):  
Neural Networks ◽  
Reinforcement Learning ◽  
Field Programmable Gate Array ◽  
Low Cost ◽  
Digital Technologies ◽  
Learning Problems ◽  
Memory Bandwidth ◽  
Q Learning ◽  
High Efficient ◽  
Field Programmable

Abstract. Due to recent advances in digital technologies, deep reinforcement learning has emerged, and has demonstrated its ability and effectiveness in solving complex learning problems not possible before. In particular, convolution neural networks (CNNs) have been demonstrated their effectiveness in reinforcement learning. However, they require intensive CPU operations and memory bandwidth that make general CPUs fail to achieve desired performance levels. In this paper, we used some low-cost field programming gates array (FPGA) designed a parallel Deep Qlearning accelerator to solve this problem. And the system has high efficient and flexibility.

scholarly journals Deep Neural Networks for Road Sign Detection and Embedded Modeling Using Oblique Aerial Images

2021 ◽  
Vol 13 (5) ◽  
pp. 879
Author(s):  
Zhu Mao ◽  
Fan Zhang ◽  
Xianfeng Huang ◽  
Xiangyang Jia ◽  
Yiping Gong ◽  
...  
Keyword(s):  
Neural Networks ◽  
Smart Cities ◽  
Geometric Constraints ◽  
Small Object ◽  
The Road ◽  
Road Signs ◽  
Road Sign ◽  
Synthesis Strategy ◽  
Urban Models ◽  
Sign Detection

Oblique photogrammetry-based three-dimensional (3D) urban models are widely used for smart cities. In 3D urban models, road signs are small but provide valuable information for navigation. However, due to the problems of sliced shape features, blurred texture and high incline angles, road signs cannot be fully reconstructed in oblique photogrammetry, even with state-of-the-art algorithms. The poor reconstruction of road signs commonly leads to less informative guidance and unsatisfactory visual appearance. In this paper, we present a pipeline for embedding road sign models based on deep convolutional neural networks (CNNs). First, we present an end-to-end balanced-learning framework for small object detection that takes advantage of the region-based CNN and a data synthesis strategy. Second, under the geometric constraints placed by the bounding boxes, we use the scale-invariant feature transform (SIFT) to extract the corresponding points on the road signs. Third, we obtain the coarse location of a single road sign by triangulating the corresponding points and refine the location via outlier removal. Least-squares fitting is then applied to the refined point cloud to fit a plane for orientation prediction. Finally, we replace the road signs with computer-aided design models in the 3D urban scene with the predicted location and orientation. The experimental results show that the proposed method achieves a high mAP in road sign detection and produces visually plausible embedded results, which demonstrates its effectiveness for road sign modeling in oblique photogrammetry-based 3D scene reconstruction.

scholarly journals Object Detection, Distributed Cloud Computing and Parallelization Techniques for Autonomous Driving Systems

2021 ◽  
Vol 11 (7) ◽  
pp. 2925
Author(s):  
Edgar Cortés Gallardo Medina ◽  
Victor Miguel Velazquez Espitia ◽  
Daniela Chípuli Silva ◽  
Sebastián Fernández Ruiz de las Cuevas ◽  
Marco Palacios Hirata ◽  
...  
Keyword(s):  
Neural Networks ◽  
Cloud Computing ◽  
Path Planning ◽  
Object Detection ◽  
Short Term Memory ◽  
Smart Cities ◽  
Autonomous Vehicle ◽  
Computational Time ◽  
End To End ◽  
On The Road

Autonomous vehicles are increasingly becoming a necessary trend towards building the smart cities of the future. Numerous proposals have been presented in recent years to tackle particular aspects of the working pipeline towards creating a functional end-to-end system, such as object detection, tracking, path planning, sentiment or intent detection, amongst others. Nevertheless, few efforts have been made to systematically compile all of these systems into a single proposal that also considers the real challenges these systems will have on the road, such as real-time computation, hardware capabilities, etc. This paper reviews the latest techniques towards creating our own end-to-end autonomous vehicle system, considering the state-of-the-art methods on object detection, and the possible incorporation of distributed systems and parallelization to deploy these methods. Our findings show that while techniques such as convolutional neural networks, recurrent neural networks, and long short-term memory can effectively handle the initial detection and path planning tasks, more efforts are required to implement cloud computing to reduce the computational time that these methods demand. Additionally, we have mapped different strategies to handle the parallelization task, both within and between the networks.

scholarly journals Real-time classification of hand movements as a basis for intuitive control of grasp neuroprostheses

2020 ◽  
Vol 6 (2) ◽  
Author(s):  
Dmitry Amelin ◽  
Ivan Potapov ◽  
Josep Cardona Audí ◽  
Andreas Kogut ◽  
Rüdiger Rupp ◽  
...  
Keyword(s):  
Neural Networks ◽  
Standard Deviation ◽  
Real Time ◽  
Convolutional Neural Networks ◽  
Recurrent Neural Networks ◽  
Healthy Subjects ◽  
Hand Movements ◽  
Cord Injury ◽  
Field Programmable

AbstractThis paper reports on the evaluation of recurrent and convolutional neural networks as real-time grasp phase classifiers for future control of neuroprostheses for people with high spinal cord injury. A field-programmable gate array has been chosen as an implementation platform due to its form factor and ability to perform parallel computations, which are specific for the selected neural networks. Three different phases of two grasp patterns and the additional open hand pattern were predicted by means of surface Electromyography (EMG) signals (i.e. Seven classes in total). Across seven healthy subjects, CNN (Convolutional Neural Networks) and RNN (Recurrent Neural Networks) had a mean accuracy of 85.23% with a standard deviation of 4.77% and 112 µs per prediction and 83.30% with a standard deviation of 4.36% and 40 µs per prediction, respectively.

Sign in / Sign up

Export Citation Format

Share Document