Nanowerk March 27, 2023 Physical reservoir computing is a computational paradigm that enables spatiotemporal pattern recognition to be performed directly in matter. The use of physical matter leads the way toward energy-efficient devices capable of solving machine learning problems without having to build a system of millions of interconnected neurons. An international team of researchers (Germany, Belgium) proposed a high-performance “skyrmion mixture reservoir” that implemented the reservoir computing model with multidimensional inputs. This implementation solved spoken digit classification tasks with an overall model accuracy of 97.4% and a < 1% word error rate. According to the researchers due to the quality of […]
Tag Archives: Pattern recognition
Image cloaking tool thwarts facial recognition programs
TechXplore August 5, 2020 To help individuals inoculate their images against unauthorized facial recognition models, researchers at the University of Chicago have developed a system called Fawkes. It helps individuals add imperceptible pixel-level changes (they call “cloaks”) to their own photos before releasing them. When used to train facial recognition models, the “cloaked” images produce functional models that consistently cause normal images of the user to be misidentified. In experiments Fawkes provided 95+% protection against user recognition regardless of how trackers train their models. They have shown that Fawkes is robust against a variety of countermeasures that try to detect […]
This object-recognition dataset stumped the world’s best computer vision models
MIT News December 10, 2019 In the real-world object detectors’ performance drops noticeably creating reliability concerns for self-driving cars and other safety-critical systems that use machine vision. A team of researchers (MIT, IBM) created ObjectNet consisting of about 50,000 photos of objects shown tipped on their side, shot at odd angles, and displayed in clutter-strewn rooms and it contains no training images. When leading object-detection models were tested on ObjectNet, their accuracy rates fell from a high of 97 percent on ImageNet to just 50-55 percent. The researchers hope that the new dataset will result in robust computer vision without […]
Preventing manipulation in automated face recognition
Fraunhofer Research October 1, 2019 In morphing processes two facial images are melded into a single synthetic facial image that contains the characteristics of both persons. As a result, biometric face recognition systems authenticate the identity of both persons based on this manipulated photo. Morphing attacks can take place before or during the process of applying for an ID document. To address this problem researchers in Germany are developing a process that identifies the image anomalies that occur during digital image processing focusing on analyzing and researching simulated imaging data using image processing, machine learning methods, and deep neural networks […]
Why did my classifier just mistake a turtle for a rifle?
MIT News July 31, 2019 We know instinctively that people and machines see the world differently, but the paper showed that the difference could be isolated and measured. Researchers at MIT have shown that a computer vision model could be compromised in a so-called black-box attack by simply feeding it progressively altered images until one caused the system to fail. Recently they highlighted multiple cases in which classifiers could be duped into confusing cats and skiers for guacamole and dogs, respectively. They trained a model to identify cats based on “robust” features recognizable to humans, and “non-robust” features that humans […]
Recovering color images from scattered light
EurekAlert July 29, 2019 Researchers at Duke University used a coded aperture, which acts as a filter that allows light to pass through some areas but not others in a specific pattern, followed by a prism. After the speckle is “stamped” by the coded aperture, it passes through a prism that causes different frequencies of light to spread out from each other. The pattern from the coded aperture shifts slightly in relation to the image being captured by the detector; the amount it shifts is directly related to the color of light passing through. They developed an algorithm that teases […]
Low-bandwidth radar technology provides improved detection of objects
Phys.org April 2, 2019 Researchers in Israel have demonstrated a ranging system which possesses superior range resolution that is almost completely free of bandwidth limitations. By sweeping over the coherence length of the transmitted signal, the partially coherent radar experimentally demonstrates an improvement of over an order of magnitude in resolving targets, compared to standard coherent radars with the same bandwidth. They developed a theoretical framework to show that the resolution could be further improved without a bound, revealing a tradeoff between bandwidth and sweep time. This concept offers solutions to problems which require high range resolution and accuracy, but […]
Engineers develop novel techniques to trick object detection systems
Science Daily April 4, 2019 To understand and document vulnerabilities in deep and machine-learning algorithms, researchers at the Southwestern Research Institute have developed patterns when worn or mounted on a vehicle, cause the algorithms in the camera to either misclassify or mislocate objects, creating a vulnerability. Malicious parties could place these patterns near roadways, potentially creating chaos for vehicles equipped with object detectors. The researchers call these patterns ‘perception invariant’ adversarial examples because they don’t need to cover the entire object or be parallel to the camera to trick the algorithm. The algorithms can misclassify the object as long as […]
Physicists train the oscillatory neural network to recognize images
Phys.org March 22, 2019 An oscillatory neural network is a complex interlacing of interacting elements that can receive and transmit oscillations of a certain frequency. Based on coupled oscillator networks implemented on vanadium dioxide structures, researchers in Russia have developed a synchronization registration method with high sensitivity and selectivity. They trained the network to synchronize only for a specific input image. In the study, the input images were transmitted to the network by changing the supply currents which changed the oscillation frequencies of oscillators. As a result, the network reacted to each received image with specific dynamics. According to the […]
Scientists teach the neural network to carry out video facial recognition — using a single photo
Eurekalert July 5, 2018 Researchers in Russia used the theory of fuzzy sets and probability theory to develop a video recognition algorithm. The algorithm significantly improves the accuracy (by 2-6% compared to earlier experiments) of identifying faces by video in real time with a small number of images for several well-known neural network architectures, such as VGGFace, VGGFace2, ResFace and LightCNN. It estimates to what degree one frame is closer to one person, and to what degree the other frame is closer to the next person. Then it compares how similar the training still photos of these two people are […]