In general, proposed solutions for LiDAR-based localization used in autonomous cars require expensive sensors and computationally expensive mapping processes. Moreover, the global localization for autonomous driving is converging to the use of maps. Straightforward strategies to reduce the costs are to produce simpler sensors and use maps already available on the Internet. Here, an analysis is presented to show how simple can a LiDAR sensor be without degrading the localization accuracy that uses road and satellite maps together to globally pose the car. Three characteristics of the sensor are evaluated: the number of range readings, the amount of noise in the LiDAR readings, and the frame rate, with the aim of finding the minimum number of LiDAR lines, the maximum acceptable noise and the sensor frame rate needed to obtain an accurate position estimation. The analysis is performed using an autonomous car in complex field scenarios equipped with a 3D LiDAR Velodyne HDL-32E. Several experiments were conducted reducing the number of frames, the number of scans per 3D point-cloud and artificially adding up to 15% of error in the ray length. Among other results, we found that using only 4 vertical lines per scan and with an artificial error added up to 15% of the ray length, the car was capable to localize itself within 2.11 meters error average. All experimental results and the followed methodology are explained in detail herein.

this work, we are proposing a computationally efficient LiDAR based lane detection system that detects both ego and side lanes using 3D LiDARs. Our solution relies on the construction of local gird map around the ego vehicle using the infrared reflectance of combination of LiDARs. To fuse the information of the LiDARs into a map, the vehicle ego-motion is taken into account. The system is built using image processing by binarizing the map to extract the lane markers. The evaluation of computational performance of the final solution is realized on a single ARM core of the NVIDIA Drive PX2 without the need for the GPUs, and achieved a frame rate of 40 Hz. In the absence of a publicly available annotated dataset for LiDAR based lane detection, we evaluate the proposed solution against our proprietary camera based lane detection system. We observed a good correlation between the two in terms of Jaccard and Dice Coefficients.

We present the Model-Predictive Motion Planner (MPMP) of the Intelligent Autonomous Robotic Automobile (IARA). IARA is a fully autonomous car that uses a path planner to compute a path from its current position to the desired destination. Using this path, the current position, a goal in the path and a map, IARA's MPMP is able to compute smooth trajectories from its current position to the goal in less than 50 ms. MPMP computes the poses of these trajectories so that they follow the path closely and, at the same time, are at a safe distance of occasional obstacles. Our experiments have shown that MPMP is able to compute trajectories that precisely follow a path produced by a Human driver (distance of 0.15 m in average) while smoothly driving IARA at speeds of up to 32.4 km/h (9 m/s).

Vehicle localization in large-scale urban environments has been commonly addressed as a map-matching problem in the literature. Generally, the maps are 2D images of the world where each pixel covers a part of it. However, building maps for large-scale urban environments requires driving the vehicle along the desired path at least once. In order to simplify this task, in this work, we propose a new localization system that uses satellite aerial map-images available on the Internet to localize a vehicle in a complex urban environment. Satellite aerial map-images are compared against re-emission maps built from the infrared reflectance information of the vehicle's LiDAR. Normalized Mutual Information (NMI) is used to compare re-emission and aerial map images. A Particle Filter Localization strategy is applied for vehicle's localization. As a result, the system has an accuracy of 0.89m in a test course with 6.5km. Our system can be used continuously without losing track, and it works even in dark and partially occluded areas.

The Virtual Generalizing Random Access Memory Weightless Neural Network (VG-RAM WNN) is an effective machine learning technique that offers simple implementation and fast training. One disadvantage of VG-RAM WNN, however, is the test time for applications with many training samples, i.e. large multi-class classification applications. In such cases, the test time tends to be high, since it increases with the size of the memory of each neuron. In this paper, we present a new methodology for handling such applications using VG-RAM WNN. By employing data clustering techniques to reduce the overall size of the neurons' memory, we were able to reduce the network's memory footprint and the system's runtime, while maintaining a high and acceptable classification performance. We evaluated the performance of our VG-RAM WNN system with compressed memory on the problem of traffic sign recognition. Our experimental results showed that, after compression, the system was able to run at very fast response times in standard computers. Also, we were able to load and run the system at interactive rates in small low-power systems, experiencing only a small reduction in classification performance.

Mapping and localization are fundamental problems in autonomous robotics. Autonomous robots need to know where they are in their area of operation to navigate through it and to perform activities of interest. In this paper, we propose an Image-Based Global Localization (VibGL) system that uses Virtual Generalizing Random Access Memory Weightless Neural Networks (VG-RAM WNN). For mapping, we employ a VG-RAM WNN that learns the world positions associated with the images captured along a trajectory. During the localization, new images from the trajectory are presented to the VG-RAM WNN, which outputs their positions in the world. We performed experiments with our VibGL system applied to the problem of localizing an autonomous car. Our experimental results show that the system is able to learn large maps (several kilometers in length) of real world environments and perform global localization with median pose precision of about 3m. Considering a tolerance of 10m VibGL is able to localize the car 95% of the time.

We present and evaluate a computer vision approach for real-time mapping of traversable road surfaces ahead of an autonomous vehicle that relies only on a stereo camera. Our system first determines the camera position with respect to the ground plane using stereo vision algorithms and probabilistic methods, and then reprojects the camera raw image to a bidimensional grid map that represents the ground plane in world coordinates. After that, it generates a road surface grid map from the bidimensional grid map using an online trained pixel classifier based on mixture of Gaussians. Finally, to build a high quality map, each road surface grid map is integrated to a probabilistic bidimensional grid map using a binary Bayes filter for estimating the occupancy probability of each grid cell. We evaluated the performance of our approach for road surface mapping in comparison to manually classified images. Our experimental results show that our approach is able to correctly map regions at 50 m ahead of an autonomous vehicle, with True Positive Rate (TPR) of 90.32% for regions between 20 and 35 m ahead and False Positive Rate (FPR) not superior to 4.23% for any range.

Virtual Generalizing Random Access Memory Weightless Neural Networks (VG-RAM WNN) is an effective machine learning technique that offers simple implementation and fast training and test. In this paper, we present a new approach for traffic sign recognition based on VG-RAM WNN. We evaluate its performance using the German Traffic Sign Recognition Benchmark (GTSRB), a large multi-class classification benchmark. Our experimental results showed that our VG-RAM WNN architecture for traffic sign recognition was able to rank at 4th position in the GTSRB evaluation system, with a recognition rate of 98.73%, and was overcome by only one automatic approach.

Virtual Generalizing Random Access Memory Weightless Neural Networks (VG-RAM WNN) is an effective machine learning technique that offers simple implementation and fast training and test. We examined the performance of VG-RAM WNN on binocular dense stereo matching using the Middlebury Stereo Datasets. Our experimental results showed that, even without tackling occlusions and discontinuities in the stereo image pairs examined, our VG-RAM WNN architecture for stereo matching was able to rank at 114th position in the Middlebury Stereo Evaluation system. This result is promising, because the difference in performance among approaches ranked in distinct positions is very small.

Several areas of knowledge are being benefited with the reduction of the computing time by using the technology of Graphics Processing Units (GPU) and the Compute Unified Device Architecture (CUDA) platform. In case of Evolutionary algorithms, which are inherently parallel, this technology may be advantageous for running experiments demanding high computing time. In this paper, we provide an implementation of the Differential Evolution (DE) algorithm in C-CUDA. The algorithm was tested on a suite of well-known benchmark optimization problems and the computing time has been compared with the same algorithm implemented in C. Results demonstrate that the computing time can significantly be reduced using C-CUDA. As far as we know, this is the first implementation of DE algorithm in C-CUDA.

With the development of Graphics Processing Units (GPU) and the Compute Unified Device Architecture (CUDA) platform, several areas of knowledge are being benefited with the reduction of the computing time. Our goal is to show how optimization algorithms inspired by Swarm Intelligence can take profit from this technology. In this paper, we provide an implementation of the Particle Swarm Optimization (PSO) algorithm in C-CUDA. The algorithm was tested on a suite of well-known benchmark optimization problems and the computing time has been compared with the same algorithm implemented in C and Matlab. Results demonstrate that the computing time can significantly be reduced using C-CUDA. As far as we know, this is the first implementation of PSO in C-CUDA.

We tackle the problem of automating the categorization of companies according to their economic activities using business descriptions in free text format as input. This categorization is vital to fundamental aspects of national governmental administration such as short, medium and long term planning and taxation. As the number of categories considered is very large (more than 1000 in the Brazilian scenario), the automatic text categorization problem targeted here is challenging. We have applied and compared the use of two different techniques to deal with it: the Vector Space Model, a well known text categorization technique; and Virtual Generalizing Random Access Memory Weightless Neural Network, or VG-RAM WNN. To our knowledge, this is the first report on using VG-RAM WNN for text categorization.