Millimeter wave radar is a promising sensing modality for enabling pervasive and privacy-preserving human sensing. However, the lack of large-scale radar datasets limits the potential of training deep learning models to achieve generalization and robustness. To close this gap, we resort to designing a software pipeline that leverages wealthy video repositories to generate synthetic radar data, but it confronts key challenges including i) multipath reflection and attenuation of radar signals among multiple humans, ii) unconvertible generated data leading to poor generality for various applications, and iii) the class-imbalance issue of videos leading to low model stability. To this end, we design Midas to generate realistic, convertible radar data from videos via two components: (i) a data generation network (DG-Net) combines several key modules, depth prediction, human mesh fitting and multi-human reflection model, to simulate the multipath reflection and attenuation of radar signals to output convertible coarse radar data, followed by a Transformer model to generate realistic radar data; (ii) a variant Siamese network (VS-Net) selects key video clips to eliminate data redundancy for addressing the class-imbalance issue. We implement and evaluate Midas with video data from various external data sources and real-world radar data, demonstrating its great advantages over the state-of-the-art approach for both activity recognition and object detection tasks.

The precision of the conventional user identification algorithm is not satisfactory because it ignores the role of user-generated data in identity matching. In this paper, we propose a frequent pattern mining-based cross-social network user identification algorithm that analyzes user-generated data in a personalized manner. We adopt the posterior probability-based information entropy weight allocation method that improves the precision rate and recall rate compared to the empirical weight allocation method. The extensive simulations are provided to demonstrate that the proposed algorithm can enhance the precision rate, recall rate, as well as the F-Measure (F1).

With the popularization of the Internet and the arrival of the big data era, numerous different social networks (SNs) have emerged to satisfy users’ social needs and offer them rich content and convenient services. Under these circumstances, identifying multiple social accounts belonging to the same user across different SNs is of great importance for many applications. Across social networks user identification (ASNUI) can help perfect user information, offer personalized service recommendation, and data mining, as well as provide support for scientific research. This paper first systematically introduces the application of ASNUI in the field of social computing, then states its applications and challenges, and reviews the adopted models, frameworks, and performance comparison state-of-the-art techniques used in ASNUI. Finally, we also identify a few future research directions in ASNUI, such as weight allocation of user attribute information, the fusion of multi-dimensional information, and large-scale user identification.

Millimeter wave radar is gaining traction recently for enabling privacy-preserving human sensing. However, the lack of large-scale, dynamic radar datasets impedes progress in developing robust and generalized deep learning models for mobile sensing applications. To address this problem, we resort to designing a software pipeline that leverages wealthy dynamic videos to generate synthetic radar data, but it faces two key challenges including i) incorrect camera and human positions leading to erroneous superposition of signal intensity and ii) the signal reflection of the background and humans in mobile scenes. To this end, we design Midas++ to utilize rich videos to generate realistic radar data via two components: (i) a human mesh fitting and calibration component calculates the camera ego-motion parameters to calibrate the extracted human positions; (ii) a reflection and noise signal estimation component combines several key modules, depth prediction , reflection model , and spatiotemporal noise estimation , to output coarse radar data, followed by a U-Net model to generate realistic radar data. We implement and evaluate Midas++ with video data from public data sources and real-world radar data, demonstrating that Midas++ outperforms other state-of-the-art approaches for both activity recognition and object detection tasks.

Propelled by advancements in deep learning technology, autonomous driving has undergone rapid development in recent years. Object sensing, as the core technology of autonomous vehicles, is undergoing a transformation from traditional single-modal sensing to multi-modal sensing, aiming to perceive environmental information in a robust, accurate, and real-time manner to adapt to increasingly complex application scenes. This paper aims to offer a comprehensive guide for radar-camera fusion, focusing on the object detection task. Firstly, detailed explanations are provided regarding the characteristics, principles, data processing methods, and representations of radar and camera data. Secondly, we analyze the evaluation metrics of object detection via a group-independent form and summarize existing radar-camera fusion datasets. Thirdly, we comprehensively review and analyze key issues in radar-camera fusion across five hierarchies (why, what, where, when, and how to fuse). Fourthly, we discuss and anticipate potential challenges and emerging research directions from the view of the robustness, accuracy, and real-time performance for object detection.

Identifying offline entities corresponding to multiple virtual accounts of users across social networks is crucial for the development of related fields, such as user recommendation system, network security, and user behavior pattern analysis. The data generated by users on multiple social networks has similarities. Thus, the concept of symmetry can be used to analyze user-generated information for user identification. In this paper, we propose a friendship networks-based user identification across social networks algorithm (FNUI), which performs the similarity of multi-hop neighbor nodes of a user to characterize the information redundancy in the friend networks fully. Subsequently, a gradient descent algorithm is used to optimize the contribution of the user’s multi-hop nodes in the user identification process. Ultimately, user identification is achieved in conjunction with the Gale–Shapley matching algorithm. Experimental results show that compared with baselines, such as friend relationship-based user identification (FRUI) and friendship learning-based user identification (FBI): (1) The contribution of single-hop neighbor nodes in the user identification process is higher than other multi-hop neighbor nodes; (2) The redundancy of information contained in multi-hop neighbor nodes has a more significant impact on user identification; (3) The precision rate, recall rate, comprehensive evaluation index (F1), and area under curve (AUC) of user identification have been improved.