Recent studies have indicated that Large Language Models (LLMs) harbor an inherent understanding of truthfulness, yet often fail to consistently express it and generate false statements. This gap between "knowing...
详细信息
In this paper, we propose an unbounded inner-product functional encryption (unbounded IPFE) scheme with semi-adaptive simulation-based security. Compared with the previous semi-adaptive secure scheme proposed by Tomid...
详细信息
In this paper, we propose an unbounded inner-product functional encryption (unbounded IPFE) scheme with semi-adaptive simulation-based security. Compared with the previous semi-adaptive secure scheme proposed by Tomida and Takashima [Asiacrypt18], our scheme enjoys about 28% shorter ciphertext and about 43% shorter secret key. Technically, we start with a bounded separable one-key IPFE scheme. In the separable one-key IPFE scheme, the public key and ciphertext can be divided into some vectors. At the same time, we develop a new transformation from a bounded separable one-key IPFE scheme towards an unbounded IPFE scheme. Finally, we give a concrete instantiation with the bounded separable one-key IPFE scheme and the transformation.
As the Internet of Things (IoT) continues to evolve, smart devices increasingly depend on edge computing for services characterized by low latency and high efficiency. The rapid proliferation of smart devices and the ...
详细信息
Few-shot learning is often challenged by low generalization performance due to the model is mostly learned with the base classes only. To mitigate the above issues, a few-shot learning method with representative globa...
详细信息
Few-shot learning is often challenged by low generalization performance due to the model is mostly learned with the base classes only. To mitigate the above issues, a few-shot learning method with representative global prototype is proposed in this paper. Specifically, to enhance generalization to novel class, we propose a strategy for jointly training base and novel classes. This process produces prototypes characterizing the class information called representative global prototypes. Additionally, to avoid the problem of data imbalance and prototype bias caused by newly added categories of sparse samples, a novel sample synthesis method is proposed for augmenting more representative samples of novel class. Finally, representative samples and non-representative samples with high uncertainty are selected to enhance the representational and discriminative abilities of the global prototype. Intensive experiments have been conducted on two popular benchmark datasets, and the experimental results show that this method significantly improves the classification ability of few-shot learning tasks and achieves state-of-the-art performance.
Event Coreference Resolution (ECR) focuses on clustering event mentions that allude to identical actual events. Previous research primarily focuses on encoding event mentions without incorporating human interpretation...
详细信息
We study a new class of MDPs that employs multinomial logit (MNL) function approximation to ensure valid probability distributions over the state space. Despite its significant benefits, incorporating the non-linear f...
Remote Memory Access (RMA) enables direct access to remote memory to achieve high performance for HPC applications. However, most modern parallel programming models lack schemes for the remote process to detect the co...
详细信息
ISBN:
(数字)9798350352917
ISBN:
(纸本)9798350352924;9798350352917
Remote Memory Access (RMA) enables direct access to remote memory to achieve high performance for HPC applications. However, most modern parallel programming models lack schemes for the remote process to detect the completion of RMA operations. Many previous works have proposed programming models and extensions to notify the communication peer, but they did not solve the multi-NIC aggregation, portability, hardware-software co-design, and usability problems. In this work, we proposed a Unified Notifiable RMA (UNR) library for HPC to address these challenges. In addition, we demonstrate the best practice of utilizing UNR within a real-world scientific application, PowerLLEL. We deployed UNR across four HPC systems, each with a different interconnect. The results show that PowerLLEL powered by UNR achieves up to a 36% acceleration on 1728 nodes of the Tianhe-Xingyi supercomputing system.
The joint classification of hyperspectral image (HSI) and light detection and ranging (LiDAR) data is gaining attention for its improved classification accuracy. However, effectively integrating the rich spectral info...
详细信息
The joint classification of hyperspectral image (HSI) and light detection and ranging (LiDAR) data is gaining attention for its improved classification accuracy. However, effectively integrating the rich spectral information of HSI and the elevation features of LiDAR has remained a challenge in multimodal fusion. This article proposes a novel approach called progressive semantic enhancement network (PSENet) for hyperspectral and LiDAR classification based on a progressive joint spatial-spectral attention mechanism. PSENet mainly comprises two modules: the spatial grouping constraint (SAGC) module and the spectral weighting constraint (SEWC) module. The SAGC module extracts multiscale features in the spatial domain, while the SEWC module focuses on enhancing semantic features in spectral dimension. By gradually utilizing spatial and spectral constraint modules to progressively enhance feature extraction, PSENet integrates affluent information for a more refined classification of ground objects. Based on experimental results, it has been demonstrated that PSENet outperforms several most advanced methods on three datasets. The SAGC and SEWC modules proposed in PSENet enable the effective integration of the spatial, spectral, and elevation information from HSI and LiDAR, providing a promising way to perform classification more accurately. The source codes of this work will be publicly available at http://***/ .
With the development of cryptographic tools such as Fully Homomorphic Encryption (FHE) and secure Multiparty Computation (MPC), privacy-preserving Machine Learning as a Service (MLaaS) has gained attractiveness for it...
详细信息
With the development of cryptographic tools such as Fully Homomorphic Encryption (FHE) and secure Multiparty Computation (MPC), privacy-preserving Machine Learning as a Service (MLaaS) has gained attractiveness for its security when it comes to utilizing cross-domain data. However, cryptographic tools are characterized by huge overhead, which results in the MLaaS quality being unbearably degraded, especially for latency-sensitive MLaaS applications. In this paper, we focus on the problem of low-latency inference associated with MLaaS and propose CrossNet, a Privacy-preserving Neural Network Inference (PPNI) framework based on FHE, for applications with limited client-side computational and communication resources. CrossNet performs model transformations on neural networks so that they can be evaluated in an FHE-friendly manner. Model transformation introduces limited interactions between client and server, thus restricting inference latency. In addition, CrossNet includes a series of layer constructions where elaborate encoding forms and computational orders are designed to further reduce the overhead of transformed layers. CrossNet outperforms the existing FHE-based frameworks by 4x efficiency and reduces nearly 30% inference latency on ResNet-50 in a resource-limited setting.
In the last few years, a myriad of Transformer based methods have drawn considerable attention due to their outstanding performance on various computer vision tasks. However, most image denoising methods are based on ...
详细信息
暂无评论