The traceback problem is widely acknowledged as one of the most challenging issues in the field of information security, and it is crucial to find effective solutions to hold attackers accountable for their actions. T...
详细信息
Recent breakthroughs in Neural Radiance Field (NeRF) based 3D reconstruction and rendering have spurred the possibility of immersive experiences in augmented and virtual reality (AR/VR). However, current NeRF accelera...
详细信息
ISBN:
(纸本)9798350350579
Recent breakthroughs in Neural Radiance Field (NeRF) based 3D reconstruction and rendering have spurred the possibility of immersive experiences in augmented and virtual reality (AR/VR). However, current NeRF acceleration techniques are still inadequate for real-world AR/VR applications due to: 1) the lack of end-to-end pipeline acceleration support, which causes impractical off-chip bandwidth demands for edge devices, and 2) limited scalability in handling large-scale scenes. To tackle these limitations, we have developed an end-to-end, scalable 3D acceleration framework called Fusion-3D, capable of instant scene reconstruction and real-time rendering. Fusion-3D achieves these goals through two key innovations: 1) an optimized end-toend processor for all three stages of the NeRF pipeline, featuring dynamic scheduling and hardware-aware sampling in the first stage, and a shared, reconfigurable pipeline with mixed-precision arithmetic in the second and third stages;2) a multi-chip architecture for handling large-scale scenes, integrating a three-level hierarchical tiling scheme that minimizes inter-chip communication and balances workloads across chips. Extensive experiments validate the effectiveness of Fusion-3D in facilitating real-time, energy-efficient 3D reconstruction and rendering. Specifically, we tape out a prototype chip in 28nm CMOS to evaluate the effectiveness of the proposed end-to-end processor. Extensive simulation based on the on-silicon measurements demonstrates a 2.5× and 6× throughput improvement in training and inference, respectively, compared to state-of-the-art accelerators. Furthermore, to assess the multi-chip architecture, we integrate four chips into a single PCB as a prototype. Further simulation results show that the multi-chip system achieves a 7.3× and 6.5× throughput improvement in training and inference, respectively, over the Nvidia 2080Ti GPU. To the best of our knowledge, Fusion-3D is the first to achieve both instant (≤ 2 second
Currently, adversarial attacks are typically designed for specific tasks, and although there are some task-agnostic attacks are generally less effective than task-specific ones. These attacks exploit the fact that CNN...
详细信息
With the profligate progress of graph-based data in numerous areas essential to observe the similarities among the graph pairs is indispensable to speed up the search outcomes. The existence of near-duplicate text doc...
详细信息
Because of the growing number of cars and the inadequate parking infrastructure, urban parking management has become more and more crucial. To predict parking spot availability in IoT-enabled intelligent systems, this...
详细信息
Recently, federated learning (FL) has emerged as a popular technique for edge AI to mine valuable knowledge in edge computing (EC) systems. To boost the performance of AI applications, large-scale models have received...
详细信息
The Integrated Health Profile (IHP) leverages blockchain technology and smart contracts to establish a decentralized and tamper-resistant platform for the storage and sharing of decentralized health records (DHRs). Se...
详细信息
ISBN:
(纸本)9798350328202
The Integrated Health Profile (IHP) leverages blockchain technology and smart contracts to establish a decentralized and tamper-resistant platform for the storage and sharing of decentralized health records (DHRs). Secured remote access to doctor-patient data is the need of the hour as the world moves towards home-based online services. In the IHP, each patient is assigned a unique identifier, and their health records are securely stored and linked to this identifier. The implementation of an individualized IHP card is the center of the IHP framework, which serves as a comprehensive repository for a patient's health records, including reports, prescriptions, medical bills, and insurance receipts. The IHP card is equipped with a unique QR code linked to the patient's unique identifier (UID). When scanned by a medical practitioner, the patient receives a one-time password (OTP) via a two-factor authentication process, providing an additional layer of consent and security. Upon successful verification, the doctor gains access to a specific subset of the patient's shared records, ensuring the patient's privacy is respected. The IHP framework significantly streamlines the customer intake process, eliminating the need for manual data entry at healthcare facility receptions. This reduction in administrative tasks minimizes waiting times and improves overall operational efficiency. Furthermore, by integrating the customer's medical history with the IHP card, medical service providers gain comprehensive insights into the customer's health status, facilitating accurate diagnoses and personalized treatment plans. To enable efficient healthcare service discovery, the IHP framework utilizes standardized application programming interfaces (APIs) to connect customer-centric applications (CCAs) with medical service provider applications (MSPs). Through these APIs, CCAs can access real-time information on available healthcare facilities, ensuring customers receive accurate and up-to
With common face databases, this study's investigation into the use of ocular traits for face detection and identification yielded encouraging findings. The principal contributions include face detection, eye dete...
详细信息
In structured light 3D measurement technology, the core challenge is recovering the phase and solving for depth from a single-frame fringe pattern. The introduction of deep learning techniques has significantly enhanc...
详细信息
In the realm of cloud computing, tackling the multi-objective task scheduling problem means finding the best way to assign tasks to virtual machines, taking into account conflicting goals like minimizing execution tim...
详细信息
暂无评论