#Reverse #Engineering of #integrated #circuits is often seen as something only companies can do, as the #equipment to image the #chip is #expensive, and the HR costs to hire enough reverse #engineers to then #understand the chip even more so. This talk gives a #short #introduction on the #motivation #behind #understanding your own or someone else’s chip (as a chip manufacturing company), and why it might be important for the rest of us (not a chip manufacturing company). The #focus is on #understanding what #millions #of #logical #gates #represent, rather than the physical aspect (delayering, imaging, image processing…), because everyone can do this at home. I will introduce some #proposed #countermeasures (like logic encryption) and #explain if, how and why they fail.
The #talk will give a #general #overview of the #research #field and #explain #why #companies #are #interested #in #reverse #engineering #ICs (IP overproduction, Counterfeits, Hardware Trojans), as well as why it’s important for an end user (IC trust, chip failure). Then, I will very shortly introduce the #reverse #engineering #workflow, from #decapsulating, #delayering, #imaging, #stitching, #image #processing and then come to the focus: #netlist #abstraction. The idea is to show some #methods which are #currently #used #in #research to #understand #what #netlists #represent. Some #theory will be #explained (#circuit #design, #formal #verification of #circuits, #graph #theory…), but I want to keep this to a #minimum. Finally, I will show some #current #ideas on #how #to #make #reverse #engineering #difficult, as well as #some #attacks on these ideas. The talk does not give insights into how large companies do reverse engineering (i.e. #throw #money #at #the #problem), but rather #show #the #research #side #of #things, with some of the methods published in the last couple of years, which is #something #everyone can do at home.
#36c3 #2019 #leipzig #randomshit
Exif metadata is not a native part of the Jpeg file format. It was an afterthought taken from the TIFF file format and tacked onto the Jpeg file format much later. This maintained backwards compatibility with old image viewers, but it meant that some programs never bothered to parse Exif data.
Most Python libraries for working with image data like numpy, scipy, TensorFlow, Keras, etc, think of themselves as scientific tools for serious people who work with generic arrays of data. They don’t concern themselves with consumer-level problems like automatic image rotation — even though basically every image in the world captured with a modern camera needs it.
This means that when you load an image with almost any Python library, you get the original, unrotated image data. And guess what happens when you try to feed a sideways or upside-down image into a face detection or object detection model? The detector fails because you gave it bad data.
You might think this problem is limited to Python scripts written by beginners and students, but that’s not the case! Even Google’s flagship Vision API demo doesn’t handle Exif orientation correctly
I’ve written about lots of computer vision and machine learning projects like object recognition systems and face recognition projects. I…medium.com