Cross-browser screen-capture detection measurements: raw logs, instrumentation, and reproduction script
收藏资源简介:
Raw measurements, instrumentation, and analysis code supporting a cross-browser study of screen-capture detection in web browsers. The dataset covers 420 staged screen captures spanning five capture methods (PrintScreen, snipping tool, screen recorder, browser extension, external camera) across three browsers (Chrome 150, Edge 151, Yandex Browser 26) on Windows 10, together with 90 legitimate control sessions. Contents:• data/E1_trials_.jsonl — per-browser detection trials with per-channel scores and state transitions• data/E2_probe_.jsonl — raw browser event probes establishing which capture signals are delivered to page scripts• data/E3_blur_samples.jsonl — labelled blur durations from a discarded timing heuristic, retained for transparency• tools/ — self-contained HTML instruments used to collect the data, re-runnable in any browser• analysis/reproduce.py — regenerates every figure reported in the accompanying manuscript Key measurements: PrintScreen detection reached 100% in Chrome and 93% in Edge, and 0% in Yandex Browser, which never delivers the PrintScreen key while delivering the snip chord. Screen recorders and browser extensions were never detected across 180 trials. No false positives occurred in 90 control sessions. All trials were performed by a single operator on a single machine. See README.md for schema documentation and replication instructions.



