English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Keystroke inference attacks are a form of side-channel attacks in which anattacker leverages various techniques to recover a user's keystrokes as sheinputs information into some display (e.g., while sending a text message orentering her pin). Typically, these attacks leverage machine learningapproaches, but assessing the realism of the threat space has lagged behind thepace of machine learning advancements, due in-part, to the challenges incurating large real-life datasets. We aim to overcome the challenge of havinglimited number of real data by introducing a video domain adaptation techniquethat is able to leverage synthetic data through supervised disentangledlearning. Specifically, for a given domain, we decompose the observed data intotwo factors of variation: Style and Content. Doing so provides four learnedrepresentations: real-life style, synthetic style, real-life content andsynthetic content. Then, we combine them into feature representations from allcombinations of style-content pairings across domains, and train a model onthese combined representations to classify the content (i.e., labels) of agiven datapoint in the style of another domain. We evaluate our method onreal-life data using a variety of metrics to quantify the amount of informationan attacker is able to recover. We show that our method prevents our model fromoverfitting to a small real-life training set, indicating that our method is aneffective form of data augmentation, thereby making keystroke inference attacksmore practical.

Leveraging Disentangled Representations to Improve Vision-Based Keystroke Inference Attacks Under Low Data