Estates, gated communities and office parks face sweeping changes to security procedures as the Information Regulator calls for access-controlled areas to collect minimal visitor information and protect privacy. The regulator has now published the “Own-Initiative Code of Conduct for Gated Access Areas” that will govern how gated-access environments handle personal information to conform to the prescripts of the Protection of Personal Information Act (Popia). The code applies to residential and commercial premises with access control. It is not a guidance note but a code of conduct, which carries much heftier weight than the former. The regulator in the code of conduct says members of the public have raised concerns that the collection of personal information at gated access entry points is excessive. “The regulator undertook research into the utilisation of closed-circuit camera (CCTV) surveillance and, in addition, considered complaints received in this regard. These collectively revealed certain access control practices of an intrusive nature, including the processing of biometric information such as the use of facial recognition systems for the purpose of positive identification of data subjects,” the draft code of conduct reads. “Furthermore, the deployment of CCTV surveillance at access control points results in the capture of facial images without the consent of data subjects and, at times, without their knowledge or awareness. “Such processing may constitute excessive collection and processing of personal information in so far as it is not relevant and limited to what is necessary for the legitimate purpose for which it is collected and accordingly warrants the imposition of appropriate regulatory measures to ensure compliance with provisions of Popia.” Under the proposed code of conduct, visitor books must not be visible to others in a queue, and digital visitor management systems must encrypt data. The code also prohibits indiscriminate copying of IDs and driver’s licences
May 11, 2026 · via businessday.co.za
New Story
107 Blog Posts To Learn About Machinelearning
by
May 10th, 2026
byLearn Repo@learn
Lets geek out. The HackerNoon library is now ranked by reading time created. Start learning by what others read most.
About Author
Lets geek out. The HackerNoon library is now ranked by reading time created. Start learning by what others read most.
Comments
TOPICS
Related Stories
109 Stories To Learn About Ocean
Oct 01, 2023
109 Stories To Learn About Ocean
Oct 01, 2023
May 10, 2026 · via hackernoon.com
New Zealand-founded retail crime intelligence company Auror has grown to more than 85,000 stores globally and increased annual revenue by 60% as retailers increasingly turn to intelligence and surveillance software to combat organised retail crime. The Auckland-founded company, which provides software used by retailers to record theft, violence and repeat offending across store networks, says much of its recent growth has come from existing customers expanding Auror into overseas markets. Watch BNZ Business Breakfast with Mei Heron live from 6am on TVNZ1 or TVNZ+ Founder and chief executive officer Phil Thomson told BNZ Business Breakfast the company's growth was increasingly being driven by existing customers taking the platform into new markets. "What's really cool is that we're being pulled by our customers - they're seeing the value in what we do, and so they're pulling us into markets that they've got stores in as well," he said. Despite the company's global growth, Thomson said Auror remained rooted in New Zealand. "While only 10% of our revenue probably comes from this market now... we overinvest, over-optimise here. It's a great place to start a company and as a test market." Auror raised $82 million in late 2024 in a Series C funding round led by US security company Axon, valuing the company at more than $500 million. Thomson said breaking into overseas markets had required rebuilding credibility from scratch. "If you look at the US, probably a great example for us... the three questions that we often got were: 'Are you a US company? Who are your US customers? And where is New Zealand?'" he said. "Even though we had great credibility down here in New Zealand and Australia, we had to start again." Auror's rapid expansion has also brought increased attention to privacy and the use of facial recognition technology.
May 10, 2026 · via 1news.co.nz
The assumption most people carry into their own minds is that the noise is the problem. That the loop of small anxieties replaying at 11pm, the way a single critical comment can surface six times before lunch, the creeping sense that something is about to go wrong even when nothing is — all of it is evidence of a mind that has turned against them. A glitch. A malfunction. Something to be quieted, medicated, or apologized for in therapy. And that assumption is almost entirely wrong. What psychology has long observed about the brain’s pattern-recognition system isn’t a story of malfunction. It’s a story of a scanner doing exactly what it was designed to do — continuously, efficiently, without ever waiting to be asked. The problem isn’t that the scanner is broken. The problem is that most people have never been told they can change what it’s scanning for. That distinction sounds small. It isn’t. - The Scanner’s Purpose: Your brain’s negativity bias isn’t broken — it’s a survival system that kept humans alive for millennia. - The Redirection Principle: You can’t turn off anxious scanning, but you can train it to notice different signals with consistent practice. - The Competence Hidden: People with overactive mental scanners often develop exceptional emotional intelligence and environmental awareness. The channel nobody chose Here is what almost nobody outside this experience understands: the brain doesn’t tune to negativity because it prefers suffering. It tunes to threat because, for most of human history, missing a threat cost more than missing a comfort. A scanner calibrated toward danger kept people alive. That same scanner, running in a body sitting at a kitchen table at 7am with a cup of tea going cold, is still faithfully doing its ancient job — flagging the unresolved email, the tone
May 10, 2026 · via leravi.org
Last week, at 4am, 19-year-old Sammy Amz was scrolling through X when something caught his eye: a popular Twitch streamer was competing in a 1v1 “mog-off” with a stranger, and losing. The next day he opened the Omoggle gaming website and began to play. Quickly he matched with another user – green dots appeared on their faces onscreen, as the website began to compare their measurements: canthal tilt, palpebral fissure ratio, nose-to-face width ratio and so on. Omoggle enables one stranger to “dominate” another in a contest of looks, which in online slang, is called mogging. It uses facial recognition to analyse and score the faces of competitors between one and 10. Omoggle’s ecosystem is based on Omegle, a now defunct site that randomly matched strangers for video-based online chats. “It’s not [scored] by looks, but it’s like, how your head is shaped, how your face is shaped,” said Amz. A week later, Amz had already competed in hundreds of mog-offs, along with some of the biggest UK streamers, emulating a trend that began in the US. On Tuesday, the Amazon-owned live-streaming platform Twitch got onboard, changing their rules to allow for “participation in current trends”, such as Omoggle. Previously, its community guidelines had prohibited the use of websites that connect a streamer to a stranger’s video feed, because of the risks of accidentally exposing its users to harmful content. To decide on a mog-off winner, Omoggle uses something called the PSL scale. The letters stand for “Perceived Sexual Market Value,” but originally, they represented three incel sites: PUAhate.com, Sluthate.com and Lookism.net. These online forums encouraged young men to develop an obsession with their physical appearance. For some it was nihilistic, and seemed to promote resentment against women who were perceived to only value physical attractiveness in men. For others,
May 10, 2026 · via theguardian.com
Hands-on building beats passive learning. Real progress comes when you train models, debug errors, and see results. Start small, finish fast. Short projects help you stay consistent and build real momentum. Reusable code accelerates learning. You spend less time setting up and more time understanding how models work. PyTorch provides a strong framework for building and training deep learning models. Project-based learning improves clarity on how systems handle data and learning steps. Real projects demonstrate how inputs move through networks and produce outputs. Practical implementation exposes training errors and their solutions. Consistent project work builds confidence and strengthens technical understanding. Let’s take a look at how these projects can shape real skills. Designed for text generation using PyTorch and poetry datasets. A neural network learns patterns in words, sentence structure, and sequence flow. The training process feeds structured text data into the engine . It predicts the next word based on earlier context. The system produces poems that match selected themes or emotions. This demonstrates how sequence models process language data. Also Read: Best 10 Tableau Projects to Try in 2026 with Source Code A tool that suggests music based on user data. Listening history is analyzed to identify patterns. User preferences are mapped to song features to improve recommendations. Suitable songs are predicted based on these patterns. The training process improves accuracy through feedback data. This shows how recommendation engines work in real platforms. Using image data to detect plant diseases. Leaf images are processed using a convolutional neural network. Visual patterns linked to diseases are identified from these inputs. Images are then classified as healthy or infected. Image preprocessing improves input quality and model performance and demonstrates practical use of computer vision in agriculture. This project connects image processing with text generation. It extracts features from images using
May 10, 2026 · via analyticsinsight.net
Sony has some of the best mirrorless offerings in the world. The company’s decision to pivot to a different format in 2012 has been their best, and it appears that it is paying off really well for many photographers. Portrait image-makers get some impressive options to choose from as well, but there is one camera that easily takes precedence over others. The camera is the Sony a1 II and features the same sensor as the a1, but it offers a new BIONZ XR. The device also has a new dedicated AI processing chip, for autofocus and image processing abilities. One also gets Pre-Capture, a burst rate of 30 fps, with full AF/AE tracking, and 8.5-stop in-body image stabilization. While the a1 II offers some amazing specs, it also has an amazing autofocus system. The a1 II has Real-time Recognition AF+ and Preset Focus, and has better tracking ability. When used with Sigma 60-600mm f4.5-6.3, it was able to track birds really well. Then with 24-105mm f4 also locks focus on with ease. It also did a fabulous job of detecting people of color in low light. When focusing on people of color in low light, this is the best Sony camera by far. Though the Nikon Z9 does a better job still and the Leica SL3 can hold its own, the Sony a1 II was able to focus on people of color very well in extremely low light. Like the autofocus, the images are pretty fantastic. The camera helps to capture details and one can use JPEGs with ease. The RAW files are versatile as always, one can retrieve with ease. The colors are nice, and one can always bump it up on RAW files. As we said in our review, “Well, the Sony a1 II is worth the upgrade
May 10, 2026 · via thephoblographer.com
When the smart battlefield turns blind: AI’s snag in the Iran war The seduction of speed The Iran war may come to be remembered as the conflict in which artificial intelligence (AI) crossed a decisive threshold, if not the Rubicon of post-modern warfare itself, moving from an auxiliary analytical tool to the pulsating nerve centre of the kill chain. What once required teams of human analysts labouring for hours over satellite imagery, drone feeds, signals intercepts, battlefield maps, and strike-option comparisons was suddenly compressed into fractions of seconds through an integrated AI-assisted architecture: Palantir’s Maven for intelligence fusion, frontier large language models (LLMs) such as Anthropic’s Claude for interpretive summarisation, and the Rapid Strike Interface (RSI) as the operational layer through which ranked outputs were rendered immediately actionable. As The Washington Post reported on March 4, 2026, Claude “generated approximately 1,000 prioritised targets on the first day of operations alone”, synthesising satellite imagery, signals intelligence, and surveillance feeds in real time and feeding these outputs into machine-structured decision sequences that drastically reduced the interval between analysis and strike authorisation. The significance of this shift lies not only in the unprecedented acceleration of data processing within contemporary military systems, but also in the transformation of the very rhythm of judgement that defines modern warfare. Across recent conflicts and advanced military simulations, AI-assisted command architectures have increasingly enabled analysts and operators to move from raw sensor inputs to synthesised battlefield interpretations at speeds previously unattainable through human-only processing. Systems such as Palantir’s Maven Smart System, deployed within US defence infrastructures, exemplify this shift towards integrated intelligence fusion, where satellite imagery, drone feeds, and signals intelligence are consolidated into unified operational displays. In some reported configurations, LLMs and generative AI tools are being explored or tested as interpretive layers to assist in summarising
May 10, 2026 · via thedailystar.net
The SCOPLE wearable AI device is an advanced piece of technology engineered to leverage advanced capabilities to help users better understand people they encounter on a daily basis. The device works by being worn and will go to work capturing information regarding a person's facial expressions, attention and more. The system will then utilize its artificial intelligence (AI) algorithm to interpret the data and provide users with information on how people react to their presence. This could work well for helping users to better understand how others perceive them or perhaps even gain more insight into conversations. The SCOPLE wearable AI device has a privacy-first design that doesn't storage any photos or videos with all data anonymized to increase peace of mind. Communication-Analyzing AI Wearables The SCOPLE Wearable AI Device Determines Emotions and More Trend Themes - Emotion-detection Wearables — Devices that infer emotional states from facial and behavioral cues enable new forms of interpersonal analytics that could reshape customer interactions and social dynamics. - Privacy-first AI Interfaces — Systems designed to anonymize and avoid storing raw sensor data are creating trust-focused AI products that balance personalization with regulatory and ethical constraints. - Real-time Social Feedback — Instant interpretation of attention and reactions during conversations is generating opportunities for context-aware augmentation of communication and presentation effectiveness. Industry Implications - Retail Customer Experience — In-store and omnichannel retailers could integrate wearable-derived emotional insights to refine merchandising, layout and personalized service strategies. - Corporate Training and HR — Learning and development programs may incorporate moment-to-moment feedback on engagement and emotional response to evaluate and tailor coaching, interviewing and leadership training. - Healthcare and Mental Health — Clinical and wellness providers stand to gain continuous, objective markers of mood and social functioning that could inform diagnosis, monitoring and treatment personalization.
May 10, 2026 · via trendhunter.com
About
Press
Copyright
Contact us
Creators
Advertise
Developers
Terms
Privacy
Policy & Safety
How YouTube works
Test new features
NFL Sunday Ticket
© 2026 Google LLC
May 10, 2026 · via youtube.com
For a deeper understanding of Israel, the region, and global antisemitism DONATE TO FATHOM Home About us Society Region Peace Anti-Zionism History Book Reviews Video Previous Issues Contributors Donate to Fathom Join over 9000 global opinion formers by subscribing to Fathom's influential Weekly Newsletter Email Address * First Name Last Name Fathom Latest Issue Spring / 2026 'Lebanon cannot afford to wait' | Fathom Interview | Hanin Ghaddar By Hanin Ghaddar On April 23, Fathom spoke to Lebanon expert Hanin Ghaddar, the Friedman Senior Fellow at the Washington Institute.... Read more > April 2026 Britain’s Sewer of Green Antisemitism By Andrew Apostolou Writing in a personal capacity, Andrew Apostolou, who is the deputy editor of Fathom and a member of... Read more > April 2026 Restoring the world’s ability to answer back: Antisemitism, Israel, and moral urgency By Joanne Strasser Joanne Strasser examines why discussions about antisemitism are so fraught. She argues that three factors prevent an actual... Read more > April 2026 A Letter to the Fourth Circle, That Is, to All of Us By Merav Roth In late March, Fathom spoke to Merav Roth (née Lapid) about her work treating trauma in Israel. She... Read more > April 2026 The Language of Loss in Gaza By Justine Hardy Justine Hardy is a mental health activist and practitioner in the field of conflict trauma, registered to practice... Read more > April 2026 Stability has a language: How the West Learned to speak Arab nationalism By Dastan Jasim Dastan Jasim argues that anti-Kurdish violence by regimes, ideologies, and political movements in Arab countries reflects a shared,... Read more > March 2026 National Conservatism’s Jewish Problem By Daniel Goldman Daniel Goldman reviews the confirmation hearing for Jeremy Carl and his views about Jews, arguing that he is... Read more > March
May 10, 2026 · via fathomjournal.org
-- On May 8, 2026, the Beijing Innovation Center of Humanoid Robotics hosted a livestream themed around the “Wise KaiWu” Agent. With the goals of being “fully autonomous, more open, and easier to use,” the event showcased breakthroughs in embodied intelligent tactile interaction. Featuring the industry’s first global scene perception and dynamic memory system, the platform enables robots to evolve from passive execution to proactive task handling, and from simple assignments to complex operations, opening up entirely new possibilities for embodied intelligence. Today, global embodied intelligence is rapidly evolving from “being able to converse” to “being able to work,” while AI Agents are moving from the digital realm into the physical world. Wise KaiWu began development a full year earlier than frameworks such as OpenClaw. After 14 months of iteration, it has achieved four major breakthroughs: spatial memory, personalized interaction at scale, one-time development with multi-robot deployment, and real-world robot validation. The result is a mass-producible and reusable professional-grade embodied intelligence solution designed to provide a practical intelligent foundation for household, commercial, and industrial applications. Spatial Memory + Personalized Intelligence: Robots That Understand Both Environments and People 1. Spatial Memory: The Industry’s First Global Dynamic Memory System Beyond “What You See Is What You Get” Traditional robots rely heavily on instantaneous visual input. Once an object leaves the field of view, it effectively “disappears,” and any environmental change can trigger “memory loss,” preventing robots from performing complex reasoning or long-duration tasks. Wise KaiWu Agent introduces the industry’s first global scene perception and dynamic spatial memory system: - Building dynamic semantic maps that record object categories, colors, positions, and spatial relationships while updating them in real time; - Enabling persistent memory across time and viewpoints, allowing robots to accurately locate items even after they leave visual range; - Supporting relational reasoning
May 10, 2026 · via markets.financialcontent.com
About
Press
Copyright
Contact us
Creators
Advertise
Developers
Terms
Privacy
Policy & Safety
How YouTube works
Test new features
NFL Sunday Ticket
© 2026 Google LLC
May 9, 2026 · via youtube.com
Disneyland and Disney California Adventure have officially transitioned to using facial recognition technology at most entry gates. This shift aims to streamline park access and enhance fraud prevention by replacing manual ticket checks with biometric validation. While Disney maintains that the system is designed for convenience, the move has reignited a broader conversation regarding data privacy and the normalization of surveillance in public spaces. How the Biometric System Works When a guest approaches a turnstile, a digital photograph is taken and processed through biometric technology. This software converts the physical features of a face into a unique numerical value. To confirm identity, the system compares this value against the image captured when the ticket or annual pass was first used. Disney’s privacy policy states that these numerical codes are typically deleted within 30 days, unless required for legal or security reasons. Industry Trends and Normalization Disney is not alone in adopting these measures. Major venues across Southern California, including the Intuit Dome and Dodger Stadium, have implemented similar programs to speed up entry. Supporters of the technology, including some park guests, suggest that facial recognition is becoming an unavoidable reality of modern life, similar to government and police surveillance. Concerns from Privacy Experts Despite the perceived convenience, legal and privacy advocates raise several critical points: - Constant Surveillance: Critics argue that because faces cannot be hidden, this technology represents a qualitative shift in surveillance where individuals are automatically identified the moment they leave home. - Data Vulnerability: Organizations like the Electronic Frontier Foundation warn that collecting biometric data creates a lucrative target for hackers. Unlike a password, biometric data cannot be changed if a breach occurs. - Accuracy and Bias: Research indicates that facial recognition systems often have higher error rates when identifying women and people of color, leading to
May 9, 2026 · via ubergizmo.com
Preserving tradition requires a new design Patterns used in ethnic clothing, products or other places are important carriers of the cultural heritage of the many ethnic groups in China. They embody rich historical memory, cultural symbolism and artistic value, vividly reflecting the diversity of Chinese civilization. However, many of these valuable pattern resources currently exist only in artifacts, images or oral traditions scattered among communities, museums and private collections. There is no systematic and digitalized integration platform and some patterns are even on the verge of disappearing due to disruptions in cultural transmission. Existing databases are mostly regional and small-scale projects without unified standards or cross-regional coordination, which greatly limits research, preservation and innovative application of these pattern resources. The unique advantages of university libraries in ethnic regions should be leveraged to establish a national database of distinctive ethnic patterns. It would support the preservation and development of traditional Chinese culture and provide foundational resources for China's cultural digitalization strategy, while also strengthening the sense of a shared Chinese national identity. First, the transformation of research outcomes remains limited. In recent years, many studies addressed ethnic patterns, some of which focused on digital technologies and database construction. Yet a relatively small number of databases have actually been completed and put into use, and most of them are small-scale platforms built by individual research groups or enterprises, with limited scope and inconsistent standards. Second, these databases are highly fragmented. Current ethnic pattern databases include industrial databases developed by enterprises, thematic databases built by universities and research teams and intangible cultural heritage databases initiated by local governments. They are largely isolated from one another without any cohesion or sharing mechanism. Third, technological applications lack sufficient cultural depth. Technologies such as AI, big data and virtual reality are increasingly used in pattern recognition,
May 9, 2026 · via chinadaily.com.cn
Summary: As AI energy demands are projected to double by 2030, researchers are developing hardware that mimics the human brain’s extreme efficiency. The study focuses on neuromorphic computing, reimagining computer architecture to process and store information simultaneously, just like biological synapses. By utilizing organic transistors, the team is laying the groundwork for AI that performs complex tasks using a fraction of the power required by conventional chips. Key Research Findings - The Efficiency Gap: While modern data centers are massive energy consumers, the human brain performs complex tasks using only about 20 watts of power. - Synaptic Architecture: Traditional chips separate memory and processing, causing energy-intensive data shuttling; Mizzou’s organic synaptic transistors perform both in the same location to eliminate this bottleneck. - The Interface Discovery: Researchers found that performance isn’t just about the material used, but the interface, the thin boundary where the semiconductor meets the insulator. - Molecular Design: Even small structural differences in materials that look identical on the surface can dramatically change how a synaptic transistor learns and adapts. - Targeted AI Tasks: This neuromorphic hardware is specifically designed to excel at pattern recognition and decision-making while consuming significantly less power. Source: University of Missouri Columbia As traditional computer chips reach their physical limits and artificial intelligence demands more energy than ever, University of Missouri researchers are rethinking how computers work by taking cues from the human brain. The timing is critical. Energy use from AI data centers is projected to double by the end of the decade, raising urgent questions about sustainability. The solution may lie in neuromorphic computing, an approach that reimagines computer hardware to process information more like biological neural networks rather than conventional chips. “One of the brain’s greatest advantages is its efficiency,” Suchi Guha, a professor of physics in Mizzou’s College
May 9, 2026 · via neurosciencenews.com
About
Press
Copyright
Contact us
Creators
Advertise
Developers
Terms
Privacy
Policy & Safety
How YouTube works
Test new features
NFL Sunday Ticket
© 2026 Google LLC
May 9, 2026 · via youtube.com
ICE’s reported smart glasses plan shows how wearable AI could move from consumer curiosity to government enforcement tool before the public has settled the rules. The next important customer for smart glasses may not be a commuter asking an AI assistant for directions. It may be a federal agent trying to identify someone in the field without pulling out a phone. That is the sharper meaning behind ICE’s reported plan to develop its own smart glasses to supplement Mobile Fortify, the facial recognition app already used by immigration officers. According to 404 Media reporting circulated through a Reddit post on Saturday, a DHS official and another person who attended a recent conference described plans for a device that would extend biometric identification into a hands-free format. The details still matter. There is a difference between a concept discussed at a conference, a prototype in procurement, and equipment deployed at scale. But the direction is clear enough to take seriously. ICE already has a mobile app that can use a phone camera for facial recognition in the field. Smart glasses would make that same workflow more ambient, quicker to trigger, and harder for bystanders to notice. For the wearable AI market, that is a meaningful signal. Consumer smart glasses are still trying to prove they can be more than a camera, speaker and chatbot wrapped around a familiar frame. Government buyers have a simpler test. Does the device give personnel faster access to identity, location, records or alerts while keeping their hands free? Startups tend to imagine smart glasses through the consumer lens first. Meta has Ray-Ban smart glasses. Apple keeps pushing spatial computing. Hardware founders talk about ambient assistants, live translation, workplace training and real-time navigation. Those are attractive markets, but they are also slow markets. Consumers care about price,
May 9, 2026 · via startupfortune.com
DeepSeek has officially launched the large-scale image recognition mode for internal testing, marking that this domestic large model has fully entered the era of multimodal interaction between text and images. After a small-scale gray-scale test in late April, DeepSeek significantly expanded access to the "Image Recognition Mode" on May 9th, and most test accounts can now access this feature through an independent entry point in the chat interface. Although the system is still labeled as "in internal testing," its layout, which is listed alongside "Quick Mode" and "Expert Mode" above the input box, indicates that multimodal understanding has become a key component of its core product matrix.
Differing from traditional simple OCR text extraction, the core of DeepSeek's latest upgrade lies in deep image recognition and semantic understanding capabilities. In practical tests, this mode can logically decompose and perceive visual information, supporting users to achieve complex cross-media interactions by directly uploading images. This move fills the gap in DeepSeek's multimodal understanding field, marking a substantial step forward in its pursuit of international top models such as GPT-4o.
May 9, 2026 · via news.aibase.com
Offering a variety of advertising and sponsorship options for reaching influential specialists from targeted demographic splits.
Cureus provides an equitable, efficient publishing and peer reviewing experience without sacrificing publication times.
Generate broad awareness and deliver relevant, peer-reviewed clinical experiences directly to potential customers.
Dedicated Cranial Radiosurgery: Clinical Experience with New & Innovative SRS Technologies
Sponsored by Zap Surgical Systems
Real-Time Adaptive Motion Management on Helical and Robotic RT Platforms
Sponsored by Accuray, Inc.
Cureus Journal of Medical Science
Please note that by doing so you agree to be added to our monthly email newsletter distribution list.
By joining Cureus, you agree to our Privacy Policy and Terms of Use.
May 9, 2026 · via cureus.com