About Press Copyright Contact us Creators Advertise Developers Terms Privacy Policy & Safety How YouTube works Test new features NFL Sunday Ticket © 2026 Google LLC
Stalled surveillance bill highlights tension between privacy and public safety Backers no longer pushing bill that would have prohibited police from using facial recognition software this session Article Summary - A bill that would stop law enforcement from using biometric surveillance failed to gain enough support for a committee vote in March and its backers aren’t going to revive it by the end of May. - Stakeholders say the bill represents a tradeoff between privacy and public safety. - But the bill only had four sponsors in the House and will have to wait at least another year for further discussion. This summary was written by the reporters and editors who worked on this story. CHICAGO — Backers of a bill aimed at limiting law enforcement’s use of biometric surveillance say they’re not looking to move the measure this legislative session. House Bill 5521, the proposed Biometric Surveillance Act, would prohibit law enforcement agencies from using or accessing facial recognition tools. But it failed to meet a March 27 committee deadline and was sent back to the House Rules committee the same day a man suspected of killing a Loyola University freshman was arrested with the help of facial recognition, according to authorities. Advocates for the bill say they are wary of debates happening in the wake of major news events, which can emphasize the technology’s role in investigations over its risks, including misidentification of individuals and expanded surveillance. Stakeholders say surveillance programs always represent a tradeoff between perceived public safety benefits and privacy concerns. “Used well, it’s a lead that can be used in conjunction with other evidence to break open a case,” said Matthew Kugler, a professor at Northwestern University Pritzker School of Law who studies how public perceptions of biometric tools shape regulation. “Used badly, it is
Mantel Group finds agentic AI is now leading customer conversations Fresh off another AWS partner award, Mantel Group is seeing a shift toward business-case AI implementations. AI is now moving away from proof-of-concept projects toward business-case implementations on agentic platforms with proper guardrails and controls. “I think over the last year it's been a genuine shift to productionising workloads,” said Andre Morgan, AWS partner lead, Mantel Group. The “factory mindset” is guiding adoption and seeing AI being used to replace business processes and accelerate software delivery life cycles (SDLC) with clear customer outcomes “It's actually achieving some results and benefits now rather than what were some of the discussions about proof of concepts a couple of years ago,” said Morgan, speaking at last week’s AWS Partner Summit in Sydney. Mantel was named AWS Consulting Partner of the Year for Australia and New Zealand in the 2026 Regional AWS Partner Awards, marking the fourth time the business has received partner recognition. Mantel’s customers include the health fund nib and Humm group as well as growing cloud adoption across regulated industries such as financial services. Its focus has been on upskilling engineers to use agentic tools for more onerous parts of the development process. At present, these applications span design, documentation, coding, and automated testing. for nib, Mantel built an AI-powered image processing pipeline that automatically extracts and pre-populates key fields from member receipt photos submitted through its mobile app, removing manual data entry. It has achieved 90 percent accuracy on key fields, made half of all claims fully touchless and delivered close to $1 million in manual handling cost savings over 12 months. “It's a good example of how we can use the technology and the customer’s business knowledge to achieve some pretty significant outcomes. It’s very high accuracy along with
Spotlight | Russia, Ukraine, United States US lawmakers call on Marco Rubio to take charge of 'stolen' Ukrainian children dossier The US intelligence community's efforts to document Russia's abuses in Ukraine are being stymied by the lack of interest shown by Director of National Intelligence Tulsi Gabbard.
As generative AI proliferates across end devices, a distributed edge-cloud architecture is taking shape — shifting inference workloads from the cloud to local hardware to reduce latency and improve efficiency. Industry players point to use cases spanning... The article requires paid subscription. Subscribe Now
Most buyers come to computer vision software expecting a search problem: find the right tool, plug it in, and watch it identify objects. The real challenge is more interesting than that. Computer vision is not a single capability. It is a family of techniques, from image classification and object detection to optical character recognition, pose estimation, and real-time anomaly detection, and different vendors have built their products around very different subsets of that family. Buying the wrong platform is not just a waste of budget. It is a six-month detour while your team rebuilds the pipeline you should have scoped from the start. Understanding What the Technology Actually Does Computer vision software teaches machines to interpret visual data: images, video streams, documents, or sensor feeds. The software applies trained models to that data and produces structured outputs, a classification label, a bounding box around an object, a count, an alert, or a confidence score. That sounds straightforward. Where it gets complicated is that "visual data" can mean a photograph, a live manufacturing camera feed, a satellite image, a PDF scan, a face captured through a webcam, or a sports broadcast frame. Each of those contexts demands different model architectures, different training datasets, and different latency requirements. A platform optimized for document processing will not perform well on real-time video. A platform built for retail shelf analysis will not translate cleanly to medical imaging. The most important question to ask before evaluating any vendor is not "can it do computer vision?" Almost all of them can. The question is: "Has this platform been designed for my specific input type, output requirement, and deployment environment?" The Main Use Cases and Why They Diverge Document and Image Analysis This is the most established segment. Software in this area extracts text and structured data
Owners of a guesthouse in north Norfolk have been left "delighted" after it was named as one of the best in the country. The White House, near Burnham Market, was listed as one of the best vineyard stays in the UK by writers at The Times, who described it as "lovely". It is owned by Alison and Mark Thompson, who renovated the property into a 12-bedroom venue for bed and breakfast visitors. The article drew attention to the guesthouse's homegrown wine, which is served at its restaurant, The Acre. Also mentioned are the wines from Cobble Hill Vineyard, just miles away from the White House. "We are delighted to have been selected, in conjunction with Cobble Hill Wines, as one of the top vineyard stays in the UK," Mrs Thompson said. "North Norfolk is such an exciting new wine destination to visit, with the amazing Cobble Hill and several other great local wineries in the area. "The local wines are very popular in Acre at the White House, our restaurant and guests often head home with a case of north Norfolk wine in their cars."
About Press Copyright Contact us Creators Advertise Developers Terms Privacy Policy & Safety How YouTube works Test new features NFL Sunday Ticket © 2026 Google LLC
New Story Can AI Solve Construction Management Bottlenecks? by May 17th, 2026 byDmytro Spilka@dmytrospilka Dmytro is the founder of Solvid and Pridicto. Featured in Hackernoon, TechRadar and Entreprepreneur. Story's Credibility About Author Dmytro is the founder of Solvid and Pridicto. Featured in Hackernoon, TechRadar and Entreprepreneur.
GO NEW MEXICO ESSENTIAL OVERLAND APP: Seeking plant identification Free iNaturalist-powered Seek app uses image recognition to help motorists and hikers name cacti, wildflowers and more A treelike cactus breaks the monotony of the desert, its branches like a candelabra reaching for the sky. At 65 mph along Interstate 40, just past Mojave National Monument in California and hundreds of miles north of Joshua Tree National Park, the scenery is a bit of a “Horse with No Name” blur. Our adventure crew wondered what this mystery plant could be. After snapping a picture and uploading it to the Seek app, we had our answer: “You observed a new species! Western Joshua Tree.” “Seek uses the power of image recognition technology to identify the plants and animals all around you,” according to the app’s description on the App store. Users can earn badges for spotting different types of plants, birds, fungi and more. Seek is to flora and fauna what Merlin is to birds. Seek shows you lists of commonly recorded insects, plants and amphibians from millions of wildlife observations vetted by human botanists and experts, according to the app description. “Our image recognition technology is based on observations submitted to iNaturalist.org Share your favorite adventure apps and New Mexico adventure channel, rigs and panorama photos with us at go@abqjournal.com. GO NEW MEXICO ADVENTURES - Not a wasteland: Chihuahuan Desert Nature Park showcases southern New Mexico desert - National parks documentary to screen at the KiMo - Fishing Line for May 14, 2026 - Water works: For summer boating New Mexicans can go to The Butte and beyond - ESSENTIAL GEAR: Shoulder-season trip tests camping gear limits A New Mexico overlander shares how a shoulder-season sprint to Sequoia and Joshua Tree led to broken equipment, improvised repairs and lessons for safer
Abstract Diabetic retinopathy (DR) is a leading cause of preventable blindness, motivating the development of reliable automated screening systems. This work proposes a Robust Deep Ensemble for Diabetic Retinopathy detection (RDE-DR) by analyzing ensemble fusion strategies. Four pre-trained convolutional neural networks (ResNet50, VGG16, VGG19, and DenseNet121) are trained using CLAHE-enhanced APTOS 2019 fundus images and integrated through seven heterogeneous fusion mechanisms, including voting-based, rank-based, and fuzzy-integral-inspired strategies. A consistent evaluation protocol is adopted, incorporating threshold optimization and probabilistic calibration analysis to validate robustness, decision margins, and accuracy–precision trade-offs. Experimental results show that multiple fusion techniques achieve comparable high performance and stable behavior on the APTOS 2019 benchmark, with the best configuration reaching 98.64% accuracy, 98.40% precision, 98.92% recall, 98.66% F1-score, and 99.78% Area-Under-Curve (AUC). Beyond peak accuracy, the study provides insights into ensemble reliability, calibration characteristics, and practical design choices for medical image classification systems. These results show that integrating transfer learning with CLAHE preprocessing and ensemble fusion yields stable experimental performance on the APTOS 2019 benchmark, suggesting potential for future medical decision support. Similar content being viewed by others Introduction Diabetic retinopathy (DR) is a leading cause of vision impairment and preventable blindness worldwide, particularly among individuals with diabetes1. Early and accurate screening is essential to prevent irreversible vision loss and improve patient outcomes2. However, early-stage DR detection remains challenging due to the subtle appearance of pathological features and the reliance on manual interpretation by clinical experts, which is time-consuming and subject to inter-observer variability3. In recent years, advances in artificial intelligence (AI) and deep learning have significantly enhanced automated disease screening, enabling reliable analysis of retinal fundus images4. Convolutional neural networks (CNNs), in particular, have demonstrated strong performance in medical image classification and segmentation tasks. Nevertheless, the scarcity of large, well-annotated medical datasets limits the training of high-capacity
Forty-three arrests after enormous £4.5m police operation keeps rival protests apart - Published The Metropolitan Police has said 43 arrests were made on Saturday during an enormous £4.5m police operation to keep demonstrators at two rival London protests apart. More than 4,000 police officers were deployed to the capital to manage a so-called "sterile zone" between a rally organised by far-right figure Tommy Robinson, and a pro-Palestinian demonstration. Tourists trying to navigate major central London landmarks were left bemused by the operation. Nelson's Column was surrounded by metal barriers. Police officers from Wales in riot uniforms prevented anyone getting from the National Gallery to Pall Mall and Regent Street. Police vehicles lined The Mall from Trafalgar Square to Buckingham Palace. On the steps above Churchill War Rooms, a two-metre-high metal barrier blocked the road past the Foreign Office towards Whitehall. But there were no serious clashes between protesters on the different marches. Every year, pro-Palestinian protesters hold a march on the weekend closest to 15 May to mark the Nakba, the displacement of Palestinians who fled or were driven from their homes during the war surrounding Israel's creation in 1948-49. But when organisers started talking to the Metropolitan Police about this year's march, they discovered 16 May had already been booked by anti-Islam campaigner Robinson, whose real name is Stephen Yaxley-Lennon, for one of his Unite the Kingdom marches. After negotiations, it was agreed that both marches could go ahead, but it meant a huge security headache for the police force. Not only were there two rival marches, but the FA Cup Final at Wembley also fell on Saturday. It meant the Met had to stage one of its most significant public order policing operations since the May Day riots of the early 2000s. The operation's aim was to maintain
OCR Studio expands KYC fraud detection for AI-generated identity documents Fake documents made with the help of generative AI are becoming increasingly more convincing. Document analysis and data extraction software firm OCR Studio has updated its document scanning tool with a new anti-fraud system that detects a wider range of AI-generated and morphed identity documents, including those made by ChatGPT, NanoBanana, Grok, Midjourney and others. The system was designed to address risks in Know Your Customer (KYC) and client onboarding processes, detecting AI-generated forgeries even when the documents appear realistic and accurate to the human eye. Instead of searching for logical inconsistencies in the document content, it examines the underlying image structure and detects low-level artifacts left by generation or editing algorithms, the company explains in a release. “Generative AI has pushed document fraud to a new level, making forged IDs more realistic and harder to spot than ever before,” points out Konstantin Bulatov, chief technology officer of OCR Studio. “The only traces left by deepfake images are invisible to the human eye and undetectable by traditional anti-fraud systems.” The rise of identity document fraud is prompting companies such as OCR Studio to develop new technologies and upgrade products. The Dubai-headquartered company presented a new neural network architecture last week that shrinks computer‑vision models by more than 40 times while maintaining accuracy. The tech could significantly improve real‑time biometric and document verification systems running on edge devices, it says. In February, the company upgraded its identity document authentication platform, OCR ID-verify system, by adding new layers of authenticity checks, allowing it to detect a broader range of document forgeries. One of the most important use cases for OCR ID-verify is online age verification. The firm has released two AI toolsets: one for in-browser age verification using ID documents, targeting compliance
POLICE used live facial recognition cameras for the first time at two rival protests yesterday — with wanted suspects and agitators arrested as they arrived for the demos. In total there were 43 arrests across the two marches in London — one a Unite the Kingdom rally organised by Tommy Robinson, the other a pro-Palestine demo. The mobile facial recognition tech was in place near King’s Cross and St Pancras railway stations — key transport hubs used by those attending the protests. Scotland Yard had vowed a “zero tolerance approach” to the whole day and said the cameras would check the profiles of passing people against a watchlist of suspects wanted for offences. The Home Office had banned 11 people described as “far-right agitators” from coming to the UK to attend the demonstrations. Right-wing figures claiming to have been barred included Polish politician Dominik Tarczynski, Belgian politician Filip Dewinter, anti-Islam commentator Valentina Gomez and Dutch activist Eva Vlaardingerbroek. READ MORE ON POLICE It was part of an “unprecedented” level of policing that had 4,000 officers on the streets to keep apart the opposing rallies, including 660 drafted in from outside the capital. The operation cost the Met £4.5million. An estimated 80,000 took part in the marches — which were held on the same day as the FA Cup Final at Wembley. As well as the 43 arrests — and 22 more at the football — four officers were assaulted and a further six were subjected to hate crime offences. Most read in The Irish Sun In response to the number of arrests, the Met said: “While this may seem high, to this point both protests have proceeded largely without significant incident.” Police helicopters and drones monitored the crowds to ensure both factions were kept apart. The Unite the Kingdom march
Telangana Deploys Facial Recognition to Automate Attendance According to the World Bank AI Repository case study, Telangana public schools use a mobile app workflow that enrolls student and teacher facial templates to automate attendance. The system captures classroom photos or selfies, matches faces to enrolled profiles, and syncs results to a central database for monitoring, per the repository. The case study identifies the deploying organisation as the Telangana school education department and the developer as RNIT Ai Solutions, and reports the solution is fully deployed and closed-source. The repository states 2.1 million facial templates are processed daily in Telangana. The technical description notes cloud-based inference, a live connection requirement, and uncertainty about whether the implementing firm fine-tuned or retrained a base computer vision model, per the World Bank entry. What happened According to the World Bank AI Repository case study, Telangana public schools have deployed a facial-recognition system to automate attendance. Per the repository, the workflow uses a mobile app where schools enroll student and teacher facial templates, then capture attendance via classroom photos (group attendance) or selfies (self-attendance) that are matched to enrolled profiles and synced to a central database. The case study lists the Telangana school education department as the deploying organisation and RNIT Ai Solutions as the developer, and records the deployment as fully deployed and closed source. The repository reports 2.1 million facial templates are processed daily in Telangana. Technical details Per the World Bank entry, the user interface is an app for teachers that connects to cloud-based inference using a computer-vision model tied to a school-specific database of uploaded face data. The case study flags uncertainty about whether the implementing firm fine-tuned or retrained an existing model versus re-using an off-the-shelf model, and notes a live network connection is required for operation. The repository classifies
Verifying your device, please wait Verification failed. Please try again. Try Again Verification failed. Please try again.
About Press Copyright Contact us Creators Advertise Developers Terms Privacy Policy & Safety How YouTube works Test new features NFL Sunday Ticket © 2026 Google LLC
Working from home has its own perils. Pets can be demanding, your back aches from hours at a desk, or you simply forget to move. There are a few apps that nudge you to move around or indicate that you’re not sitting in an ideal position, but they’re easy to dismiss. I’ve spent the better part of a decade at a home desk, iterating on the setup as I go — gaming chair, lumbar support, the works. None of it guarantees good posture. Then I came across Isa, a desk device from German startup Deep Care that takes a different approach entirely. It tracks posture, hydration, light, sound, and movement. And it does all of it without a camera or an internet connection, which, in an era of always-on surveillance, is a meaningful differentiator. Here’s how it works and what’s inside. Isa has a 5.5-inch IPS HD screen and looks like a table clock. It is powered by USB-C; the company supplies a power unit with it, but you can use any of your existing chargers too, as it has a power consumption rating of roughly 2.45W. The key sensor for the device is the Time-of-Flight (ToF) 3D depth sensor on the front — the same technology used in facial recognition and some smartphone cameras — that tracks posture and movement. It also enables beta features, such as counting the number of times you’ve had water or other liquids. The company said that the sensor works in the range of 0.15 meters to 1.8 meters. That means if the device is sitting on your desk, it can measure your movement, even when you stand up and move about. It also packs several other sensors: a ToF 1D sensor, a gyroscope, a barometer, a light sensor, a sound level sensor, a
Explore the things you love . Log into Facebook Email or mobile number Password Log in Forgot password? Create new account English (US) Español Français (France) 中文(简体) العربية Português (Brasil) Italiano More languages… Sign Up Log In Messenger Facebook Lite Video Meta Pay Meta Store Meta Quest Ray-Ban Meta Meta AI Instagram Threads Privacy Policy Consumer Health Privacy Privacy Center About Create ad Create Page Developers Careers Cookies Ad choices Terms Help Contact Uploading & Non-Users Meta © 2026
Police say 11 people arrested 'for a variety of offences' The Met said it has arrested 11 people “for a variety of offences” so far. It did not specify how many arrests were linked to the Unite the Kingdom event and the pro-Palestine march. Police said earlier today two men were arrested near Euston station on suspicion of grievous bodily harm following an incident in Birmingham where a man was run over. They arrived in London to attend the United the Kingdom protest. Organisers of the pro-Palestine Nakba Day rally estimated at least a quarter of a million people have attended, making it “10 times bigger” than the Unite the Kingdom demonstration, they said. Numbers for the Tommy Robinson march appear to be significantly down on his last show of strength in September, according to early police estimates. The Metropolitan police believe that 60,000 attended the far right Unite the Kingdom march. It is markedly down on the 150,000 the Met estimated at the same protest in September. A total of 31 arrests have been made at the Unite the Kingdom and Nakba Day protests in London as of about 4.30pm on Saturday, police said. Nick Tenconi, the head of Nigel Farage’s former party Ukip, was spotted at the far-right march. Tenconi has said he wants to use the military to “round up and deport the Islamists, illegals and the communists”. The Nakba protest has concluded and all participants have left the area, the Metropolitan Police said. The force added Parliament Square and Whitehall are now largely clear of protesters, but the road will remain at least partially closed while the infrastructure is removed. Numbers for the Tommy Robinson march appear to be significantly down on his last show of strength in September, according to early police estimates. The Metropolitan