Traffic Sign Classification

The traffic signs engraved on the streets nowadays improve traffic security by advising the driver regarding speed limits or any further potential perils like profound thrilling streets, inescapable fix street works or any common intersections. With the quick improvement of economy and innovation in the cutting edge society, vehicles have become an imperative method for transportation in the day by day travel of individuals. Albeit the fame of autos has acquainted impressive comfort with individuals, it has additionally caused a various traffic security issues that can't be overlooked, for example, gridlock and successive street mishaps. Traffic security issues are to a great extent brought about by abstract reasons identified with the driver, like obliviousness, inappropriate driving activity and resistance with traffic rules, and keen vehicles have become a compelling way to wipe out these human components. Self-driving innovation can help, or even autonomously complete the driving activity, which is vital to free the human body and extensively lessen the rate of mishaps. Traffic sign identification and acknowledgment are significant in the advancement of astute vehicles, which straightforwardly influences the execution of driving practices. Traffic sign identification and grouping is of vital significance for the fate of independent vehicle innovation. We benchmark the commented on dataset with AI baselines Convolutional Neural Organizations (CNN). Computational strategies for AI (ML) have shown their importance for the projection of possible outcomes for educated choices. AI calculations have been applied for quite a while in numerous applications. An information driven methodology with higher precision as here can be extremely valuable for a proactive reaction from the public authority and residents. At long last, we propose a bunch of exploration openings and arrangement justification for additional useful applications. Keywords: Convolutional Neural Networks, Traffic sign detection, Traffic safety, Computational Methods, machine Learning Algorithms

Traffic Sign Classification Using CNN

With the increasing necessity of autonomous electric vehicles as time passes by, there are a lot of technical prospects within the structure that have a lot of scope for advancement. One such prospect is traffic sign recognition. Several models have already been developed and are in practice but it is evident to everyone within this field that there is still a lot of untapped potential. In this project we implement feature classification using convolutional neural networks to achieve an efficiency and accuracy higher than that of a conventional model. The system first converts the image into grayscale and then three layers of the image are created. By using skilled convolutional neural network which incorporates crucial data of traffic signs and images, they are parallelly assigned to corresponding classes. Results have shown that this system works with great efficiency. Index Terms: Traffic sign, GTSRB, ALVINN, Neural Network, CNN, LeNet, ReLU, Evaluation, Epoch

A collection of easily deployable adversarial traffic sign stickers

The emergence of new autonomous driving systems and functions – in particular, systems that base their decisions on the output of machine learning subsystems responsible for environment perception – brings a significant change in the risks to the safety and security of transportation. These kinds of Advanced Driver Assistance Systems are vulnerable to new types of malicious attacks, and their properties are often not well understood. This paper demonstrates the theoretical and practical possibility of deliberate physical adversarial attacks against deep learning perception systems in general, with a focus on safety-critical driver assistance applications such as traffic sign classification in particular. Our newly developed traffic sign stickers are different from other similar methods insofar that they require no special knowledge or precision in their creation and deployment, thus they present a realistic and severe threat to traffic safety and security. In this paper we preemptively point out the dangers and easily exploitable weaknesses that current and future systems are bound to face.

Traffic Sign Classification and Detection of Indian Traffic Signs using Deep Learning

This paper presents an effective solution to detecting traffic signs on road by first classifying the traffic sign images us-ing Convolutional Neural Network (CNN) on the German Traffic Sign Recognition Benchmark (GTSRB)[1] and then detecting the images of Indian Traffic Signs using the Indian Dataset which will be used as testing dataset while building classification model. Therefore this system helps electric cars or self driving cars to recognise the traffic signs efficiently and correctly. The system involves two parts, detection of traffic signs from the environment and classification based on CNN thereby recognising the traffic sign. The classification involves building a CNN model of different filters of dimensions 3 × 3, 5 × 5, 9 × 9, 13 × 13, 15 × 15,19 × 19, 23 × 23, 25 × 25 and 31 ×31 from which the most efficient filter is chosen for further classifying the image detected. The detection involves detecting the traffic sign using YOLO v3-v4 and BLOB detection. Transfer Learning is used for using the trained model for detecting Indian traffic sign images.

Traffic Sign Classification Using Convolutional Neural Networks and Computer Vision

The world is quickly and continuously advancing towards better technological advancements that will make life quite easier for us, human beings [22]. Humans are looking for more interactive and advanced ways to improve their learning. One such dream is making a machine think like a computer, which lead to innovations like AI and deep learning [25]. The world is running at a higher pace in the domain of AI, deep learning, robotics and machine learning Using this knowledge and technology, we could develop anything right now [36]. As a part of sub-domain, the introduction of Convolution Neural Networks made deep learning extensively strong in the domain of image classification and detection [1]..The research that we have conducted is one of its kind. Our research used Convolution Neural Network, TensorFlow and Keras.

Exploration of Hardware Acceleration Methods for an XNOR Traffic Signs Classifier

Deep learning algorithms are a key component of many state-of-the-art vision systems, especially as Convolutional Neural Networks (CNN) outperform most solutions in the sense of accuracy. To apply such algorithms in real-time applications, one has to address the challenges of memory and computational complexity. To deal with the first issue, we use networks with reduced precision, specifically a binary neural network (also known as XNOR). To satisfy the computational requirements, we propose to use highly parallel and low-power FPGA devices. In this work, we explore the possibility of accelerating XNOR networks for traffic sign classification. The trained binary networks are implemented on the ZCU 104 development board, equipped with a Zynq UltraScale+ MPSoC device using two different approaches. Firstly, we propose a custom HDL accelerator for XNOR networks, which enables the inference with almost 450 fps. Even better results are obtained with the second method - the Xilinx FINN accelerator - enabling to process input images with around 550 frame rate. Both approaches provide over 96% accuracy on the test set.

Exploration of Hardware Acceleration Methods for an XNOR Traffic Signs Classifier

Deep learning algorithms are a key component of many state-of-the-art vision systems, especially as Convolutional Neural Networks (CNN) outperform most solutions in the sense of accuracy. To apply such algorithms in real-time applications, one has to address the challenges of memory and computational complexity. To deal with the first issue, we use networks with reduced precision, specifically a binary neural network (also known as XNOR). To satisfy the computational requirements, we propose to use highly parallel and low-power FPGA devices. In this work, we explore the possibility of accelerating XNOR networks for traffic sign classification. The trained binary networks are implemented on the ZCU 104 development board, equipped with a Zynq UltraScale+ MPSoC device using two different approaches. Firstly, we propose a custom HDL accelerator for XNOR networks, which enables the inference with almost 450 fps. Even better results are obtained with the second method - the Xilinx FINN accelerator - enabling to process input images with around 550 frame rate. Both approaches provide over 96% accuracy on the test set.

Learning Region-Based Attention Network for Traffic Sign Recognition

Traffic sign recognition in poor environments has always been a challenge in self-driving. Although a few works have achieved good results in the field of traffic sign recognition, there is currently a lack of traffic sign benchmarks containing many complex factors and a robust network. In this paper, we propose an ice environment traffic sign recognition benchmark (ITSRB) and detection benchmark (ITSDB), marked in the COCO2017 format. The benchmarks include 5806 images with 43,290 traffic sign instances with different climate, light, time, and occlusion conditions. Second, we tested the robustness of the Libra-RCNN and HRNetv2p on the ITSDB compared with Faster-RCNN. The Libra-RCNN performed well and proved that our ITSDB dataset did increase the challenge in this task. Third, we propose an attention network based on high-resolution traffic sign classification (PFANet), and conduct ablation research on the design parallel fusion attention module. Experiments show that our representation reached 93.57% accuracy in ITSRB, and performed as well as the newest and most effective networks in the German traffic sign recognition dataset (GTSRB).