Search for: All records

With rich visual data, such as images, becoming readily associated with items, visually-aware recommendation systems (VARS) have been widely used in different applications. Recent studies have shown that VARS are vulnerable to item-image adversarial attacks, which add human-imperceptible perturbations to the clean images associated with those items. Attacks on VARS pose new security challenges to a wide range of applications, such as e-commerce and social media, where VARS are widely used. How to secure VARS from such adversarial attacks becomes a critical problem. Currently, there is still a lack of systematic studies on how to design defense strategies against visual attacks on VARS. In this article, we attempt to fill this gap by proposing anadversarial image denoising and detectionframework to secure VARS. Our proposed method can simultaneously (1) secure VARS from adversarial attacks characterized bylocalperturbations by image denoising based onglobalvision transformers; and (2) accurately detect adversarial examples using a novel contrastive learning approach. Meanwhile, our framework is designed to be used as both a filter and a detector so that they can bejointlytrained to improve the flexibility of our defense strategy to a variety of attacks and VARS models. Our approach is uniquely tailored for VARS, addressing the distinct challenges in scenarios where adversarial attacks can differ across industries, for instance, causing misclassification in e-commerce or misrepresentation in real estate. We have conducted extensive experimental studies with two popular attack methods (FGSM and PGD). Our experimental results on two real-world datasets show that our defense strategy against visual attacks is effective and outperforms existing methods on different attacks. Moreover, our method demonstrates high accuracy in detecting adversarial examples, complementing its robustness across various types of adversarial attacks. 

We developed a new method for simultaneously visualizing RNAs and proteins in plant cells. This method works well for endogenous as well as infectious RNAs and their cognate binding proteins. 

ABSTRACT Viroids are single‐stranded circular noncoding RNAs that mainly infect crops. Upon infection, nuclear‐replicating viroids engage host DNA‐dependent RNA polymerase II for RNA‐templated transcription, which is facilitated by a host protein TFIIIA‐7ZF. The sense‐strand and minus‐strand RNA intermediates are differentially localised to the nucleolus and nucleoplasm regions, respectively. The factors and function underlying the differential localisation of viroid RNAs have not been fully elucidated. The sense‐strand RNA intermediates are cleaved into linear monomers by a yet‐to‐be‐identified RNase III‐type enzyme and ligated to form circular RNA progeny by DNA ligase I (LIG1). The subcellular compartment for the ligation reaction has not been characterised. Here, we show that LIG1 and potato spindle tuber viroid (PSTVd) colocalise near the nucleolar region inNicotiana benthamianaprotoplasts. The colocalised region is also the highly condensed region of sense‐strand PSTVd RNA, indicating that PSTVd RNA and LIG1 form a biomolecular condensate for RNA processing. This finding expands the function of biomolecular condensates to the infection of subviral pathogens. In addition, this knowledge of viroid biogenesis will contribute to exploring thousands of viroid‐like RNAs that have been recently identified. 

Abstract RNA comprises a versatile group of biomolecules that play diverse roles in a wide range of biological processes. From synthesis to degradation, RNAs interact with cognate proteins that assist in processes such as transcription, splicing, modification, trafficking, and the execution of their functions. While numerous valuable techniques exist to study RNA-protein interactions, observing RNAs and their associated proteins simultaneously within cells remains a challenge, despite its potential to provide deeper insights into RNA-protein interactions. In this study, we adapted a modified immunofluorescence (IF) assay combined with RNA fluorescence in situ hybridization (FISH) to successfully visualize the colocalization of potato spindle tuber viroid with RNA polymerase II in the nucleus. This new method that combines IF and FISH will facilitate future studies on RNA and protein colocalization in various plant systems.