In this Letter, we demonstrate high-precision end-to-end adaptive optics (AO) technique based on the X-Shape Fusion Transformer-convolutional neural network (XFTC-Net) without an additional probe path to compensate fo...
详细信息
As an important part of the new generation of information technology,the Internet of Things(IoT)has been widely concerned and regarded as an enabling technology of the next generation of health care *** fundus photogr...
详细信息
As an important part of the new generation of information technology,the Internet of Things(IoT)has been widely concerned and regarded as an enabling technology of the next generation of health care *** fundus photography equipment is connected to the cloud platform through the IoT,so as to realize the realtime uploading of fundus images and the rapid issuance of diagnostic suggestions by artificial *** the same time,important security and privacy issues have *** data uploaded to the cloud platform involves more personal attributes,health status and medical application data of *** leaked,abused or improperly disclosed,personal information security will be ***,it is important to address the security and privacy issues of massive medical and healthcare equipment connecting to the infrastructure of IoT healthcare and health *** meet this challenge,we propose MIA-UNet,a multi-scale iterative aggregation U-network,which aims to achieve accurate and efficient retinal vessel segmentation for ophthalmic auxiliary diagnosis while ensuring that the network has low computational complexity to adapt to mobile *** this way,users do not need to upload the data to the cloud platform,and can analyze and process the fundus images on their own mobile terminals,thus eliminating the leakage of personal ***,the interconnection between encoder and decoder,as well as the internal connection between decoder subnetworks in classic U-Net are redefined and ***,we propose a hybrid loss function to smooth the gradient and deal with the imbalance between foreground and *** with the UNet,the segmentation performance of the proposed network is significantly improved on the premise that the number of parameters is only increased by 2%.When applied to three publicly available datasets:DRIVE,STARE and CHASE DB1,the proposed network achieves the accuracy/F1-score of 96.33%/84.34%
Retrieval-augmented generation (RAG) is considered to be a promising approach to alleviate the hallucination issue of large language models (LLMs), and it has received widespread attention from researchers recently. D...
详细信息
ISBN:
(纸本)9798400704314
Retrieval-augmented generation (RAG) is considered to be a promising approach to alleviate the hallucination issue of large language models (LLMs), and it has received widespread attention from researchers recently. Due to the limitation in the semantic understanding of retrieval models, the success of RAG heavily lies on the ability of LLMs to identify passages with utility. Recent efforts have explored the ability of LLMs to assess the relevance of passages in retrieval, but there has been limited work on evaluating the utility of passages in supporting question answering. In this work, we conduct a comprehensive study about the capabilities of large language models (LLMs) in utility evaluation for open-domain question answering (QA). Specifically, we introduce a benchmarking procedure and collection of candidate passages with different characteristics, facilitating a series of experiments with five representative LLMs. Our experiments reveal that: (i) well-instructed LLMs can distinguish between relevance and utility, and that LLMs are highly receptive to newly generated counterfactual passages. Moreover, (ii) we scrutinize key factors that affect utility judgments in the instruction design. And finally, (iii) to verify the efficacy of utility judgments in practical retrieval augmentation applications, we delve into LLMs' QA capabilities using the evidence judged with utility and direct dense retrieval results. (iv) We propose a.. -sampling, listwise approach to reduce the dependency of LLMs on the sequence of input passages, thereby facilitating subsequent answer generation. We believe that the way we formalize and study the problem along with our findings contributes to a critical assessment of retrieval-augmented LLMs. Our code and benchmark can be found at https://***/ict-bigdatalab/utility_judgments.
Owing to the huge volume of big data, users generally use the cloud to store big data. However, because the data are out of the control of users, sensitive data need to be protected. The ciphertext-policy attribute-ba...
详细信息
Owing to the huge volume of big data, users generally use the cloud to store big data. However, because the data are out of the control of users, sensitive data need to be protected. The ciphertext-policy attribute-based encryption scheme can not only effectively control the access of big data, but also decrypt the ciphertext as long as the user's attributes satisfy the access structure of ciphertext, so as to realize one to many big data sharing. When the user's attributes do not satisfy the access structure of ciphertext, the attribute-based proxy re-encryption scheme can be used for big data sharing. The ciphertext-policy attribute-based proxy re-encryption (CP-ABPRE) scheme combines the characteristics of the ciphertext-policy attribute-based encryption scheme and proxy re-encryption scheme. In a CP-ABPRE scheme, on the one hand, the data owner can use the ciphertext-policy attribute-based encryption scheme to encrypt the big data for cloud storage, to realize the access control of the big data. On the other hand, the proxy (cloud service provider) can convert ciphertext under one access structure into ciphertext under another access structure, thus realizing big data sharing between users of different attribute sets. In this article, we modify the existing attribute-based encryption scheme based on Ring Learning With Errors (RLWE), add re-encryption key generation algorithm, re-encryption ciphertext generation algorithm, and re-encryption ciphertext decryption algorithm, and construct CP-ABPRE scheme. In the construction of the re-encryption key, we introduce a random vector and hide the vector in the key by threshold technology. Finally, a CP-ABPRE scheme supporting threshold access structure is constructed based on RLWE. Compared with the existing attribute-based proxy re-encryption schemes, our scheme has smaller public parameters, can encrypt multiple plaintext bits at a time, and can resist selective access structure and chosen plaintext attack, so it is m
Nowadays, companies use databases for information storage and it is inevitable that there will be confidential information which can cause severe crises and public relations problems if leaked. In order to solve this ...
详细信息
Gaze tracking technology has a wide range of applications in the fields of VR/AR glasses. In this paper, a plastic optical fiber (POF) is used as a light transmission waveguide to transmit the light reflected from the...
详细信息
Gaze tracking technology has a wide range of applications in the fields of VR/AR glasses. In this paper, a plastic optical fiber (POF) is used as a light transmission waveguide to transmit the light reflected from the eye to the outside of the device for processing, thereby eliminating the need for in-device camera installations. By processing two fan-shaped surfaces on the POF at a 45 degrees angle relative to the vertical direction, the POF gains the ability to couple light from the side. The reflected light of the eye can be transmitted to the outside of the device through the POF. The specklegram corresponding to 76 different gaze directions is classified by SE-Resnet18, and the accuracy reached 96.9%. The gaze tracking system is low cost and simple in structure and has potential application in fields such as AR glasses, human-computer interaction (HCI), and medical diagnosis. (c) 2024 Optica Publishing Group. All rights, including for textand data mining (TDM), Artificial Intelligence (AI) training, and similar technologies, are reserved.
A uniform experimental design(UED)is an extremely used powerful and efficient methodology for designing experiments with high-dimensional inputs,limited resources and unknown underlying models.A UED enjoys the followi...
详细信息
A uniform experimental design(UED)is an extremely used powerful and efficient methodology for designing experiments with high-dimensional inputs,limited resources and unknown underlying models.A UED enjoys the following two significant advantages:(i)It is a robust design,since it does not require to specify a model before experimenters conduct their experiments;and(ii)it provides uniformly scatter design points in the experimental domain,thus it gives a good representation of this domain with fewer experimental trials(runs).Many real-life experiments involve hundreds or thousands of active factors and thus large UEDs are *** large UEDs using the existing techniques is an NP-hard problem,an extremely time-consuming heuristic search process and a satisfactory result is not *** paper presents a new effective and easy technique,adjusted Gray map technique(AGMT),for constructing(nearly)UEDs with large numbers of four-level factors and runs by converting designs with s two-level factors and n runs to(nearly)UEDs with 2^(t−1)s four-level factors and 2tn runs for any t≥0 using two simple transformation *** justifications for the uniformity of the resulting four-level designs are given,which provide some necessary and/or sufficient conditions for obtaining(nearly)uniform four-level *** results show that the AGMT is much easier and better than the existing widely used techniques and it can be effectively used to simply generate new recommended large(nearly)UEDs with four-level factors.
Public-key Encryption with keyword Search (PEKS) enables secure keyword searches within encrypted data. At the same time, Public-key Authenticated Encryption with keyword Search (PAEKS) enhances security by permitting...
详细信息
Public-key Encryption with keyword Search (PEKS) enables secure keyword searches within encrypted data. At the same time, Public-key Authenticated Encryption with keyword Search (PAEKS) enhances security by permitting authorized users to search specific keyword sets, protecting against Internal keyword Guessing Attacks (IKGA). However, to the best of our knowledge, existing PEKS and PAEKS schemes typically require to generate a distinct set of keyword ciphertext for each data user, leading to storage, computation, and communication costs and the lack of support for multiple-keyword search. In this article, we introduce a novel, efficient public-key searchable encryption scheme from the private set intersection (PSI) with scalable proxy servers, using a PSI protocol with multiple proxy server settings, which achieves sub-linear complexity. Our scheme is secure against IKGA and supports multiple keyword searches and sharing one encrypted keyword set by multiple users. We introduce an efficient system model with scalable proxy servers, significantly reducing computational overhead through a divide-and-conquer approach. Our proposed scheme supports multiple data users, and multiple keyword searches, utilizing a single set of keyword ciphertext for multiple data users. We formally define a security model and present a comprehensive security proof to demonstrate that our scheme maintains ciphertext-indistinguishability and trapdoor-indistinguishability.
Password managers (PMs) provide users with convenient and robust functionalities to manage their credentials, highly recommended by security experts and major standard bodies. One of the most popular features is the a...
详细信息
ISBN:
(纸本)9798331520885
Password managers (PMs) provide users with convenient and robust functionalities to manage their credentials, highly recommended by security experts and major standard bodies. One of the most popular features is the autofill functionality, with which users need a single click or a few clicks to fill in every field in web forms, facilitating the process of completing web forms. However, such indiscriminate autofill brings severe privacy threats. PMs may inadvertently fill data into wrong fields in web forms, even hidden fields, potentially leading to privacy leaks and credential theft. In this paper, we conduct an empirical study evaluating the effectiveness of 30 popular PMs in identifying and handling hidden fields. We focus on the privacy threats posed by the autofill functionality, which fills data into hidden fields. We develop a semi-automated autofill testing tool and explore whether PMs autofill sensitive data into hidden fields across 15 concealment techniques and three web forms, including personal information, credit card, and login forms. Experimental results reveal that every PM autofills data into hidden fields in at least one web form, with an overall filled probability of 58.7% in 1032 scenarios. Further analysis reveals that login forms are the most vulnerable, with a 65.7% probability of hidden fields autofill. Hidden fields concealed by clip-path and content-visibility are filled with passwords by all PMs. Besides, built-in-browser PMs exhibit a 4.07 times higher likelihood of filling data into hidden fields than separately-installed PMs. Even more concerning, built-in-browser PMs, except Safari, autofill passwords into hidden fields under any concealment technique. 37.7% of autofill scenarios with insufficient user interaction pose heightened privacy threats, as users are unaware of autofill content. These privacy threats have been confirmed by popular PMs like LastPass. To mitigate the threats brought by the autofill functionality, we present two
Information retrieval aims to find information that meets users' needs from the corpus. Different needs correspond to different IR tasks such as document retrieval, open-domain question answering, retrieval-based ...
详细信息
Information retrieval aims to find information that meets users' needs from the corpus. Different needs correspond to different IR tasks such as document retrieval, open-domain question answering, retrieval-based dialogue, and so on, while they share the same schema to estimate the relationship between texts. It indicates that a good IR model can generalize to different tasks and domains. However, previous studies indicate that state-of-the-art neural information retrieval (NIR) models, e.g., pre-trained language models (PLMs) are hard to generalize. It is mainly because the end-to-end fine-tuning paradigm makes the model overemphasize task-specific signals and domain biases but loses the ability to capture generalized essential signals. To address this problem, we propose a novel NIR training framework named NIR-Prompt for retrieval and reranking stages based on the idea of decoupling signal capturing and combination. NIR-Prompt exploits Essential Matching Module (EMM) to capture the essential matching signals and gets the description of tasks by Matching Description Module (MDM). The description is used as task-adaptation information to combine the essential matching signals to adapt to different tasks. Experiments under in-domain multi-task, out-of-domain multi-task, and new task adaptation settings show that NIR-Prompt can improve the generalization of PLMs in NIR for both retrieval and reranking stages compared with baselines.
暂无评论