TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Sensor Fingerprints: Camera Identification and Beyond

49 点作者 haltingproblem将近 2 年前

5 条评论

buildbot将近 2 年前
Interestingly, reading the fourth chapter, the first time this technique of identifying the sensor based on the specific noise characteristics was 2005. I bet this was tacitly well known far before 2005.<p>There are many camera from pre - 2005 that did dark frames and flat frames, for example Sinar digital backs. They all came with a CD with the unique calibration file for the back to tune out these characteristics to a large degree. Without them, you can see literally see the difference in exposure down the center of the image from the two stitched lithography mask shots required to make large CCDs at the time.
评论 #36161683 未加载
muhehe将近 2 年前
Does this survive jpeg (or similar) compression or do you need raw file for this?
评论 #36164548 未加载
neodypsis将近 2 年前
A problem I see is that someone could denoise an image and then add adversarial noise so that it resembles one of these signatures.
评论 #36163826 未加载
tokai将近 2 年前
Would taping a minor radiation source to the camera body be enough to scramble the noise fingerprint?
keskival将近 2 年前
If you can recognize such a fingerprint, you can also trivially fake it or wipe it with an optimizer in the loop with the recognizer.