Adversarial attacks are not something new to the world of Deep Networks used for image recognition. However, as the research with Deep Learning grows, more flaws are uncovered. The team at the University of KU Leuven in Belgium have demonstrated how, by simple using a colored photo held near the torso of a man can render him invisible to image recognition systems based on convolutional neural networks.
Convolutional Neural Networks or CNNs are a class of Deep learning networks that reduces the number of computations to be performed by creating hierarchical patterns from simpler and smaller networks. They are becoming the norm for image recognition applications and are being used in the field. In this new paper, the addition of color patches is seen to confuse the image detector YoLo(v2) by adding noise that disrupts the calculations of the CNN. The patch is not random and can be identified using the process defined in the publication.
This attack can be implemented by printing the disruptive pattern on a t-shirt making them invisible to surveillance system detection. You can read the paper[PDF] that outlines the generation of the adversarial patch. Image recognition camouflage that works on Google’s Inception has been documented in the past and we hope to see more such hacks in the future. Its a new world out there where you hacking is colorful as ever.
Continue reading “The Cloak Of Invisibility Against Image Recognition”
Often it feels as if soldering is deemed to be more of an art form than something that’s underpinned by the cold, hard reality of physics and chemistry. From organic chemistry with rosin, to the material properties of fragile gold bond wires and silicon dies inside IC packages and the effects of thermal stress on the different parts of an IC package, it’s a complicated topic that deserves a lot more attention than it usually gets.
A casual inquiry around one’s friends, acquaintances, colleagues and perfect strangers on the internet usually reveals the same pattern: people have picked up a soldering iron at some point, and either figured out what seemed to work through trial and error, or learned from someone else who has learned what seemed to work through trial and error. Can we say something scientific about soldering?
Continue reading “Get To Know The Physics Behind Soldering And The Packaging Of ICs”
Classic games never seem to have gone out of style and with the emulation powers of the Raspberry Pi, there seems to be no end of projects folks have been coming up with. [Chris Mills] project is a great looking monitor to get his Commodore 64 fix by combining the retro looks of a home-made 64-style monitor with the Raspberry Pi.
[Chris] is only interested in Commodore 64 emulation, at least with this project, and wanted something that would fit on a desk without taking up too much room. An eight inch LCD security monitor fit the bill perfectly. [Chris] ended up building a wooden enclosure for the monitor to give it that Commodore look. The monitor, power supply and cable connections fit inside along with speakers; each of these having their inputs on the back. A fan vents in the back as well and the Pi sits outside running the Combian 64 emulation software.
[Chris] has put up some galleries of build pics. The logo from the old Commodore logo is a nice touch. Read over the Hackaday site and you could build your own Commodore 64, or use the Commodore 64 itself to house the Raspberry Pi if you wanted.
When you think of Fortran you probably think of punched cards and green bar paper. While it is true that Fortran isn’t the go-to language it used to be — pun unintentional — it still has a vibrant community of people who do serious number crunching. However, many members of that community have been seduced away by interactive tools that are also good at number crunching like MATLAB, Julian, and Python with special libraries. The LFortran project aims to create a Fortran environment with interactivity like Python, but retaining the speed that Fortran is known for.
The resulting tool is impressive. You can use it from Jupyter, can parse code targeting existing Fortran compilers, and supports Linux, Mac, and Windows. There is development to make the code fully interoperable with other languages like C or Python as well as take advantage of GPUs and other specialized hardware. They are also zeroing in on full Fortran 2018 support.
Continue reading “Fortran Goes Interactive”