Lifeboat Foundation

Chrome is rolling out an update for Android users that lets them lock their incognito sessions with a password code or biometric info when they exit the app. The feature has been available for iOS users for some time, but now it’s being made available to folks using Chrome on Android.

Users can activate this feature by going to Chrome Settings Privacy & Security and turning on the “Lock incognito tabs when you close Chrome” toggle. So next time when a user exits Chrome, their incognito session will automatically be locked. To unlock the incognito tabs, you can use the biometric unlock on the phone such as a fingerprint unlock or lock code.

This post is also available in: עברית (Hebrew)

Fire departments conducting “size up” training typically rely on whiteboard discussions, drives around neighborhoods and photo-based systems. New training technology will help firefighters train for different types of fires or hazardous material situations, vehicle accidents, residential and commercial buildings, etc. An augmented reality training tool for firefighters, called Forge, uses artificial intelligence and biometric training to simulate real emergencies. Developed by Avrio Analytics, the system is designed to make sure that firefighters possess communication, situational awareness and associated skills needed in emergencies.

“Biometric and performance data collected during training allows Forge’s AI to dynamically change the training based on the user’s cognitive load, such as providing more or less guidance to the individual or introducing new training complexity in real-time,” the company told govtech.com. “This allows for training sessions tailored to the ability of the individual.”

When a German security researcher, Matthias Marx, found a United States military device for sale on eBay—an instrument previously used to identify wanted individuals and known terrorists during the War in Afghanistan—Marx gambled a little and placed a low bid of $68.

He probably didn’t expect to win, since he offered less than half the seller’s asking price, $149.95. But win he did, and after that, he had an even bigger surprise coming, The New York Times reported. When the device arrived with a memory card still inside, Marx was shocked to realize he had unwittingly purchased the names, nationalities, photographs, fingerprints, and iris scans of 2,632 people whose biometric data had allegedly been scanned by US military.

The device allegedly stored not just personal identifiable information (PII) of seemingly suspicious persons, but also of US military members, people in Afghanistan who worked with the government, and ordinary people temporarily detained at military checkpoints. Most of the data came from residents of Afghanistan and Iraq.

Face recognition tools are computational models that can identify specific people in images, as well as CCTV or video footage. These tools are already being used in a wide range of real-world settings, for instance aiding law enforcement and border control agents in their criminal investigations and surveillance efforts, and for authentication and biometric applications. While most existing models perform remarkably well, there may still be much room for improvement.

Researchers at Queen Mary University of London have recently created a new and promising architecture for face recognition. This architecture, presented in a paper pre-published on arXiv, is based on a strategy to extract facial features from images that differs from most of those proposed so far.

“Holistic methods using convolutional neural networks (CNNs) and margin-based losses have dominated research on face recognition,” Zhonglin Sun and Georgios Tzimiropoulos, the two researchers who carried out the study, told TechXplore.

Check out all the on-demand sessions from the Intelligent Security Summit here.

The adoption of a password-free future is hyped by some of the biggest tech companies, with Apple, Google, and Microsoft committing to support the FIDO standard this past May. Along with the Digital ID Bill reintroduced to Congress this past July, we’re poised to take a giant leap away from the password to a seemingly more secure digital future. But as we approach a post-password world, we still have a long way to go in ensuring the security of our digital lives.

As companies continue developing solutions to bridge us to a passwordless world, many have prioritized convenience over security. Methods of two-factor authentication (2FA) and multi-factor authentication (MFA) such as SMS or email verification — or even the use of biometrics — have emerged as leading alternatives to the traditional username/password. But here’s the catch: Most of these companies are validating devices alone and aren’t properly leveraging this technology, leaving the door open for bad actors.

Researchers at Universidad Autónoma de Madrid have recently created an innovative, AI-powered platform that could enhance remote learning, allowing educators to securely monitor students and verify that they are attending compulsory online classes or exams.

An initial prototype of this platform, called Demo-edBB, is set to be presented at the AAAI-23 Conference on Artificial Intelligence in February 2022, in Washington, and a version of the paper is available on the arXiv preprint server.

“Our investigation group, the BiDA-Lab at Universidad Autónoma de Madrid, has substantial experience with biometric signals and systems, behavior analysis and AI applications, with over 300 hundred published papers in last two decades,” Roberto Daza Garcia, one of the researchers who carried out the study, told TechXplore.

Computer scientists at New York University and Michigan State University have trained an artificial neural network to create fake digital fingerprints that can bypass locks on cell phones. The fakes are called “DeepMasterPrints”, and they present a significant security flaw for any device relying on this type of biometric data authentication. After exploiting the weaknesses inherent in the ergonomic needs of cellular devices, DeepMasterPrints were able to imitate over 70% of the fingerprints in a testing database.

An artificial neural network is a type of artificial intelligence comprising computer algorithms modeled after the human brain’s ability to recognize patterns. The DeepMasterPrints system was trained to analyze sets of fingerprint images and generate a new image based on the features that occurred most frequently. This “skeleton key” could then be used to exploit the way cell phones authenticate user fingerprints.

In cell phones, the necessarily small size of fingerprint readers creates a weakness in the way they verify a print. In general, phone sensors only capture a partial image of a print when a user is attempting to unlock the device, and that piece is then compared to the phone’s authorized print image database. Since a partial print means there are fewer characteristics to distinguish it than a full print, a DeepMasterPrint needs to match fewer features to imitate a fingerprint. It’s worth noting that the concept of exploiting this flaw is not unique to this particular study; however, generating unique images rather than using actual or synthesized images is a new development.

Amazon is bringing its palm print-scanning biometric payment technology to several Whole Foods locations.

Biometrics: Every person has measurable physical characteristics that are unique to them — and because these attributes are unique and measurable, they can be used to verify our identity.

Biometric technologies — like the one that probably unlocks your phone — automate this verification, analyzing a face, fingerprint, or palm for distinct identifiers linked to a specific person.

The number of AI and, in particular, machine learning (ML) publications related to medical imaging has increased dramatically in recent years. A current PubMed search using the Mesh keywords “artificial intelligence” and “radiology” yielded 5,369 papers in 2021, more than five times the results found in 2011. ML models are constantly being developed to improve healthcare efficiency and outcomes, from classification to semantic segmentation, object detection, and image generation. Numerous published reports in diagnostic radiology, for example, indicate that ML models have the capability to perform as good as or even better than medical experts in specific tasks, such as anomaly detection and pathology screening.

It is thus undeniable that, when used correctly, AI can assist radiologists and drastically reduce their labor. Despite the growing interest in developing ML models for medical imaging, significant challenges can limit such models’ practical applications or even predispose them to substantial bias. Data scarcity and data imbalance are two of these challenges. On the one hand, medical imaging datasets are frequently much more minor than natural photograph datasets such as ImageNet, and pooling institutional datasets or making them public may be impossible due to patient privacy concerns. On the other hand, even the medical imaging datasets that data scientists have access to could be more balanced.

In other words, the volume of medical imaging data for patients with specific pathologies is significantly lower than for patients with common pathologies or healthy people. Using insufficiently large or imbalanced datasets to train or evaluate a machine learning model may result in systemic biases in model performance. Synthetic image generation is one of the primary strategies to combat data scarcity and data imbalance, in addition to the public release of deidentified medical imaging datasets and the endorsement of strategies such as federated learning, enabling machine learning (ML) model development on multi-institutional datasets without data sharing.