Place recognition is an important technique for autonomous cars to achieve full autonomy since it can provide an initial guess to online localization algorithms. Although current methods based on images or point cloud...
详细信息
ISBN:
(纸本)9781665491907
Place recognition is an important technique for autonomous cars to achieve full autonomy since it can provide an initial guess to online localization algorithms. Although current methods based on images or point clouds have achieved satisfactory performance, localizing the images on a large-scale point cloud map remains a fairly unexplored problem. This cross-modal matching task is challenging due to the difficulty in extracting consistent descriptors from images and point clouds. In this paper, we propose the I2P-Rec method to solve the problem by transforming the cross-modal data into the same modality. Specifically, we leverage on the recent success of depth estimation networks to recover point clouds from images. We then project the point clouds into Bird's Eye View (BEV) images. Using the BEV image as an intermediate representation, we extract global features with a Convolutional Neural Network followed by a NetVLAD layer to perform matching. The experimental results evaluated on the KITTI dataset show that, with only a small set of training data, I2P-Rec achieves recall rates at Top-1% over 80% and 90%, when localizing monocular and stereo images on point cloud maps, respectively. We further evaluate I2P-Rec on a 1 km trajectory dataset collected by an autonomous logistics car and show that I2P-Rec can generalize well to previously unseen environments.
This paper presents an innovative approach to automatic volume control using imageprocessing and deep learning techniques. The ability to automatically adjust volume levels based on environmental factors and user pre...
详细信息
ISBN:
(数字)9798350375237
ISBN:
(纸本)9798350375244
This paper presents an innovative approach to automatic volume control using imageprocessing and deep learning techniques. The ability to automatically adjust volume levels based on environmental factors and user preferences has significant implications for various audio applications, including teleconferencing systems, smart devices, and public address systems. By combining imageprocessingalgorithms with deep learning models, this paper aims to develop a robust and adaptive volume control system capable of accurately adjusting audio levels in real-time. The paper discusses the theoretical foundations, technical implementation, experimental results, and potential applications of the proposed automatic volume control system.
A pothole is a significant disadvantage in the roadways, resulting from depressions formed due to the improper quality of road materials and external conditions, such as extreme weather forces and the heavy vehicles p...
详细信息
The precise boundaries of the cadastral parcels from the Unmanned Aerial Vehicle (UAV) data are essential for any eGovernance application. The pix2pix, image-to-image translation using the conditional Generative Adver...
详细信息
The precise boundaries of the cadastral parcels from the Unmanned Aerial Vehicle (UAV) data are essential for any eGovernance application. The pix2pix, image-to-image translation using the conditional Generative Adversarial Network (cGAN) models, has emerged as an alternative to the traditional machine learning and imageprocessingalgorithms. It has been used and demonstrated for productive purposes in different domains without any change in the pix2pix network model and loss functions. The pix2pix model is implemented in this research for extracting the cadastral parcel boundaries using the existing UAV data set, and the corresponding digitised data. The input data set is prepared using the python modules. The model is also used to predict the synthetic UAV data from the map data. The predicted boundary of the model is very useful. The proposed model can reduce the manual labour and human interventions in outlining the parcel boundary from UAV data.
Despite their success, the traditional deep neural networks are unable to rapidly generalize from a limited number of samples. To address the issue of large training data requirements with deep neural networks, one-sh...
详细信息
To detect surface damage to buildings, it is necessary to involve workers who are at risk of industrial injuries when inspecting hard-to-reach areas of industrial premises. Attraction of special means, such as aerial ...
详细信息
The application of CBCT systems in intraoperative environments has become increasingly common, but concurrent CBCT systems are unsuitable for situations that require a large longitudinal imaging FoV, such as orthopedi...
详细信息
This research work proposes an efficient framework to improve multispectral image classification through convolutional neural networks (CNNs) with optimized hyperspectral band selection. Hyperspectral images contain e...
详细信息
Visual pollution is a significant obstacle in the modern era, where the world is advancing towards increasingly diverse inventions. These inventions require a suitable environment to achieve accurate outcomes. Artific...
详细信息
ISBN:
(纸本)9783031664304;9783031664311
Visual pollution is a significant obstacle in the modern era, where the world is advancing towards increasingly diverse inventions. These inventions require a suitable environment to achieve accurate outcomes. Artificial intelligence has already permeated all fields and interests of life;similarly, visual pollution also needs to be addressed properly. Visual pollution often creates obstacles in performing various tasks. To mitigate these issues, an artificial intelligence-based model will play a vital role. This work deals with detecting visual pollution using an artificial intelligence-based algorithm to apply practical solutions that enhance urban public scenery. In the first step, a dataset is chosen from an authorized organization;specifically, the data is sourced from Mendeley, named the Saudi Arabia Public Roads Visual Pollution Dataset 2023. The second step involves data scaling and background removal from training images to facilitate learning in AI models. In the third step, the dataset is processed using Random Forest and support vector machine algorithms to visualize the model's accuracy results. The support vector machine demonstrates better performance compared to the Random Forest.
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processing s...
ISBN:
(纸本)9783893180950
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processingsystems;data coverage for detecting representation bias in image datasets: a crowdsourcing approach;balancing utility and fairness in submodular maximization;stateful entities: object-oriented cloud applications as distributed dataflows;learning over sets for databases;a new PET for data collection via forms with data minimization, full accuracy and informed consent;adaptive compression for databases;analysis of open government datasets from a data design and integration perspective;fine-grained geo-obfuscation to protect workers’ location privacy in time-sensitive spatial crowdsourcing;and a framework to evaluate early time-series classification algorithms.
暂无评论