Video question answering (VideoQA), whose goal is to produce answers through the integration of linguistic and visual understanding, has emerged as a significant research focus. Although Large Multimodal Models (LMMs) and autonomous agent methods have achieved notable advances in VideoQA, excessive computational overhead and restricted multimodal interaction capabilities limit their ability to facilitate the continuous evolution of the VideoQA system. To address the challenge, we introduce DigimonGPT, an evolvable VideoQA agent inspired by cognitive psychology. Specifically, DigimonGPT integrates a multimodal memory mechanism to achieve the continuous evolution of VideoQA systems. An intra-video declarative memory contains fundamental features of the video and semantic contexts extracted from historical QA pairs. Another inter-task procedural memory encodes task-solving experience for further question answering. Additionally, we introduce a hierarchical memory replay mechanism for VideoQA that selects appropriate memories by their relevance and question complexity. Extensive experiments demonstrate that DigimonGPT's accuracy averagely outperforms 13.71% on NExT-QA datasets and 9.89% on Intent-QA datasets over LMM and autonomous agents.

Video object detection stands as a pivotal element within the burgeoning landscape of autonomous driving systems. The exigency to fulfill stringent real-time requisites, while upholding both precision and efficiency in detection, underscores its significance. Although extant methodologies enhance either accuracy or efficiency through the exploitation of spatio-temporal inter-dependencies within the video context, their propensity to conduct detection on discrete frames begets superfluous computations and curbed real-time efficacy. This paper introduces a pioneering approach, called Real-Batch, tailored explicitly to redress this quandary. Real-Batch ingeniously processes batches of video frames uniformly, effectually winnowing out repetitive object detection occurrences. Our methodology is rigorously evaluated on the real-word datasets, scrutinizing four key metrics: accuracy, efficiency, informational value, and adherence to timing constraints. The comprehensive findings substantiate that Real-Batch yields an unparalleled maximal surge in accuracy and efficiency, increasing of 4.2%-13.2% and 24.7%-43.9%, respectively, offering promising advancements for autonomous driving systems.

Artificial intelligence (AI) model inference performance in browsers is constrained, and transmitting data to the server consumes substantial transfer time by cloud computing. In this paper, we investigate the status quo of cloud and browser processing and explore model computation partitioning methods. Our study is rooted in WebGPU and employs the Tensorflow.js framework, encompassing seven AI models spanning computer vision, natural language processing, and automatic speech recognition domains. Leveraging the characteristics of neural network layers, we find a significant performance boost through a method that partitions AI models at layer granularity. We design a system called WebInf to partition AI models at layer granularity between the browser and server for faster inferencing-based adaptive model partitioning. WebInf supports diverse hardware, wireless networks, neural network structures, servers, and adaptive partitioning models for optimal inference performance. We evaluate WebInf on two laptops and servers, demonstrating that WebInf yields inference time improvements of 30% and 52%, respectively, when compared to separate inference execution in servers and browsers. The improvements can even peak at 33% and 69% respectively.

With the rapid development of the Internet of Things (IoT) and the rapid popularization of 5 G networks, the data that needs to be processed in Mobile Crowdsourcing (MCS) system is increasing every day. Traditional cloud computing can no longer meet the needs of crowdsourcing for real-time data and processing efficiency, thus, edge computing was born. Edge computing can be calculated at the edge of network so that greatly improve the efficiency and real-time performance of data processing. In addition, most of the existing privacy protection technologies are based on the trusted third parties. Therefore, in view of the semi-trustworthiness of edge servers and the transparency of blockchain, this paper proposes a triple real-time trajectory privacy protection mechanism (T-LGEB) based on edge computing and blockchain. Through combining the localized differential privacy and multiple probability extension mechanism, the T-LGEB mechanism is proposed to send the requests and data to the edge server in this paper. Then, through the spatio-temporal dynamic pseudonym mechanism proposed in the paper, the entire trajectory of task participants is divided into multiple unrelated trajectory segments with different pseudonymous identities in order to protect the trajectory privacy of task participants while ensuring high data availability and real-time data. Through a large number of experiments and comparative analysis on multiple real data sets, the proposed T-LGEB has extremely high privacy protection capabilities and data availability, and the resource consumption caused is relatively low.

With the rise of the Internet of Things (IoT) and fifth-generation (5G) networks, which have led to a surge in data processing and increased data transfer time, traditional cloud computing could no longer meet the needs of workers, so edge computing has emerged. Edge computing could meet the demand for low time consumption by processing data at the edge of the network and then transmitting it to a third-party platform. However, since the credibility of the third-party platform is unknown which can easily leak the privacy of workers. For the transparent mechanism of blockchain, a two-stage privacy protection mechanism based on blockchain is proposed to solve this problem. In the first stage, this paper proposes a double disturbance localized differential privacy (DDLDP) algorithm to disturb the location information of workers. In the second stage, all the sensing data are uploaded to the blockchain through edge nodes, processed by the edge cloud, and fed back to the requester. Blockchain technology not only guarantees the integrity of sensing data, but also prevents the possibility of third-party platforms from leaking workers' privacy. Through extensive performance evaluation and comparative experiments on real data sets, the DDLDP algorithm could effectively protect the privacy of workers and has higher service quality and data availability.

In the participatory sensing framework, privacy protection of the Internet of Things (IoT) is very important. In this article, cryptography-based methods are utilized to protect participants' privacy information in unsecured network channels for dynamic and real-time sensing tasks. The edge computing paradigm is introduced in the traditional participatory sensing framework to reduce network latency. Then, the Rivest Cipher 4 stream cipher and logistic mapping are combined to deal with the problems of participants' limited resources and untruthful third-party platforms. Finally, the product algebra and logistic mapping are combined to deal with the problems of large numbers of participants' access and poor randomness of keystream. Through extensive performance evaluation and comparison experiments on the real-world data, the effectiveness and adaptation of the proposed privacy protection based on stream cipher are verified. It could effectively solve the problem of poor network latency and improve the privacy protection level of IoT.

In spatiotemporal crowdsourcing applications, sensing data uploaded by participants usually contain spatiotemporal sensitive data. If application servers publish the unprocessed sensing data directly, it is easy to expose the privacy of participants. In addition, application servers usually adopt the static publishing mechanism, which is easy to produce problems such as poor timeliness and large information loss for spatiotemporal crowdsourcing applications. Therefore, this paper proposes a spatiotemporal privacy protection (STPP) method based on dynamic clustering methods to solve the privacy protection problem for crowd participants in spatiotemporal crowdsourcing systems. Firstly, the working principles of a dynamic privacy protection mechanism are introduced. Then, based on k-anonymity and l-diversity, the spatiotemporal sensitive data are anonymized. In addition, this paper designs the dynamic k-anonymity algorithm based on the previous anonymous results. Through extensive performance evaluation on real-world data, compared with existing methods, the proposed STPP algorithm could effectively solve the problem of poor timeliness and improve the privacy protection level while reducing the information loss of sensing data.