Results 31 to 40 of about 15,386,676 (302)
Boosting Backdoor Attack with A Learnable Poisoning Sample Selection Strategy [PDF]
arXiv, 2023 Data-poisoning based backdoor attacks aim to insert backdoor into models by manipulating training datasets without controlling the training process of the target model. Existing attack methods mainly focus on designing triggers or fusion strategies between triggers and benign samples.
arxiv
DeepPoison: Feature Transfer Based Stealthy Poisoning Attack [PDF]
arXiv, 2021 Deep neural networks are susceptible to poisoning attacks by purposely polluted training data with specific triggers. As existing episodes mainly focused on attack success rate with patch-based samples, defense algorithms can easily detect these poisoning samples.
arxiv
Poisoning and Backdooring Contrastive Learning [PDF]
arXiv, 2021 Multimodal contrastive learning methods like CLIP train on noisy and uncurated training datasets. This is cheaper than labeling datasets manually, and even improves out-of-distribution robustness. We show that this practice makes backdoor and poisoning attacks a significant threat.
arxiv
BMC Public Health, 2021 Background In China, long working hours and insomnia are relatively common among internet company employees. Considering that both can affect mental health, we examined their independent and interaction effects on these employees’ depressive symptoms (DS)
Xiaoman Liu +4 more
doaj +1 more source
Towards One Shot Search Space Poisoning in Neural Architecture Search [PDF]
arXiv, 2021 We evaluate the robustness of a Neural Architecture Search (NAS) algorithm known as Efficient NAS (ENAS) against data agnostic poisoning attacks on the original search space with carefully designed ineffective operations. We empirically demonstrate how our one shot search space poisoning approach exploits design flaws in the ENAS controller to degrade ...
arxiv
Benzodiazepine poisoning in elderly [PDF]
Vojnosanitetski Pregled, 2016 Background/Aim. Benzodiazepines are among the most frequently ingested drugs in self-poisonings. Elderly may be at greater risk compared with younger individuals due to impaired metabolism and increased sensitivity to benzodiazepines. The aim of
Perković-Vukčević Nataša +4 more
doaj +1 more source
Universal Backdoor Attacks [PDF]
arXiv, 2023 Web-scraped datasets are vulnerable to data poisoning, which can be used for backdooring deep image classifiers during training. Since training on large datasets is expensive, a model is trained once and re-used many times. Unlike adversarial examples, backdoor attacks often target specific classes rather than any class learned by the model.
arxiv
Clinical toxicology, 2017 Introduction: This is the 34th Annual Report of the American Association of Poison Control Centers’ (AAPCC) National Poison Data System (NPDS). As of 1 January 2016, 55 of the nation’s poison centers (PCs) uploaded case data automatically to NPDS.
D. Gummin +5 more
semanticscholar +1 more source
Analysis and Detectability of Offline Data Poisoning Attacks on Linear Dynamical Systems [PDF]
arXiv, 2022 In recent years, there has been a growing interest in the effects of data poisoning attacks on data-driven control methods. Poisoning attacks are well-known to the Machine Learning community, which, however, make use of assumptions, such as cross-sample independence, that in general do not hold for linear dynamical systems.
arxiv
Trends of drug-related poisoning cases attended to at a poison control center.
Revista brasileira de epidemiologia = Brazilian journal of epidemiology, 2019 INTRODUCTION Drug-related poisonings have a high impact on morbidity and mortality, representing the first cause of intoxication in Brazil. OBJECTIVE To describe the trends of cases of drug-related poisonings attended to by a poison control center ...
Thays Lopes Mathias +2 more
semanticscholar +1 more source