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.
You can unsubscribe anytime.
By joining Cureus, you agree to our Privacy Policy and Terms of Use.
Apr 6, 2026 · via cureus.com
Microsoft announced three new in-house AI models, MAI-Transcribe-1, MAI-Voice-1, and MAI-Image-2, expanding its capabilities across speech recognition, voice generation, and image creation. The models are now available through Microsoft Foundry and the MAI Playground, with the company emphasizing improved performance, speed, and cost efficiency. MAI-Transcribe-1 is designed for speech-to-text transcription and delivers state-of-the-art performance across the top 25 most-used languages based on the FLEURS benchmark. The model is optimized for real-world environments and delivers batch transcription speeds 2.5 times faster than Microsoft’s previous Azure Fast offering. Microsoft also highlighted its strong price-to-performance positioning compared to other large cloud providers. MAI-Voice-1 focuses on generating natural, expressive speech, preserving speaker identity across long-form content while enabling developers to create custom voices with only a few seconds of audio. The model can generate up to 60 seconds of audio per second and is built for high GPU efficiency, supporting scalable deployment for enterprise use cases. It is already being integrated into Copilot experiences, including audio-based features and podcasts. MAI-Image-2 enhances Microsoft’s image generation capabilities, delivering faster performance while maintaining high-quality outputs suitable for professional creative workflows. The model has ranked among the top three on the Arena.ai leaderboard and is being rolled out across Microsoft products, including Bing and PowerPoint. Early enterprise adoption includes WPP, which is using the model for large-scale creative production. Microsoft is positioning these models as “better, faster, and cheaper” than competing offerings, with aggressive pricing aimed at developers and enterprise customers. Pricing starts at $0.36 per hour for MAI-Transcribe-1, $22 per million characters for MAI-Voice-1, and $5 per million tokens for text input and $33 per million tokens for image output for MAI-Image-2. The company also emphasized its broader vision of “humanist AI,” focusing on building models that are aligned with human needs and designed for safe, responsible
Apr 5, 2026 · via pulse2.com
Artificial Intelligence (AI) is transforming the modern world at an incredible pace, creating powerful opportunities across industries. From healthcare to finance, entertainment to education, AI is driving innovation and reshaping how businesses operate. If you are planning your future, exploring the Top Careers in Artificial Intelligence can help you secure a high-demand, future-proof profession. In this article, we will explore the Top AI Careers, their roles, required skills, and why they are in such high demand. Whether you are a student or a professional looking to switch careers, understanding the Top Careers in Artificial Intelligence can guide you toward a successful path. Machine Learning Engineer One of the most in-demand roles among the Top Careers in Artificial Intelligence is a Machine Learning Engineer. These professionals design algorithms that allow machines to learn from data and improve over time without being explicitly programmed. To excel in this field, you need strong programming skills (Python, R), knowledge of statistics, and experience with frameworks like TensorFlow or PyTorch. Machine Learning Engineers are at the core of many Top AI Careers due to their role in building intelligent systems. Data Scientist Data Scientists are key players in the Top Careers in Artificial Intelligence landscape. They analyze large datasets to extract insights and help businesses make data-driven decisions. This role requires expertise in data analysis, statistics, machine learning, and visualization tools. The increasing reliance on data makes this one of the most important Top AI Careers today. AI Research Scientist AI Research Scientists focus on advancing the field of artificial intelligence. They work on cutting-edge technologies such as deep learning, natural language processing, and computer vision. This is one of the most advanced Top Careers in Artificial Intelligence, often requiring a PhD in computer science or a related field. These professionals push the boundaries of
Apr 5, 2026 · via academiamag.com
AI can identify mushrooms in seconds by analyzing a photo and matching it against massive datasets using advanced computer vision. Powered by models like Convolutional Neural Networks, the system detects patterns such as cap shape, color, gills, and texture to predict the species instantly. Instead of relying on bulky field guides, users simply upload an image. The AI scans features, compares them with thousands of labeled samples, and delivers accurate results within seconds. This makes mushroom identification faster, safer, and accessible—even for complete beginners exploring nature. AI mushroom identification uses intelligent algorithms to recognize fungi from images. It blends data science with visual recognition to simplify a traditionally complex process. Rather than memorizing species, users rely on trained systems that “learn” from millions of images and improve over time. Together, these technologies simulate expert-level identification—but in seconds. Manual identification is a skill that takes years to master, especially in Mycology. Many mushrooms look nearly identical but differ significantly in safety, making the process risky for beginners. The genus Amanita contains both visually appealing and highly toxic mushrooms. Misidentification here can have serious consequences. AI reduces these risks by offering quick, data-backed insights. AI follows a structured process to ensure fast and reliable results. The user takes or uploads a clear photo of the mushroom. The system scans: The AI compares extracted features with a trained dataset of thousands of mushroom images. The system returns: This entire process takes just a few seconds. AI-powered tools offer advantages that traditional methods cannot match. These benefits make AI tools ideal for both casual users and serious foragers. AI is transforming how people interact with nature and fungi. Imagine walking through a forest and spotting an unfamiliar mushroom. Instead of guessing, you scan it with an AI tool and instantly learn its identity and
Apr 5, 2026 · via ocnjdaily.com
New Story
The Hidden Audio Bias Inside Audio-Visual Speech Recognition
by
April 5th, 2026
byaimodels44@aimodels44
Among other things, launching AIModels.fyi ... Find the right AI model for your project - https://aimodels.fyi
About Author
Among other things, launching AIModels.fyi ... Find the right AI model for your project - https://aimodels.fyi
Apr 5, 2026 · via hackernoon.com
I write about tech every day, and taking screenshots in Windows is a big part of my work. And for the longest time, I relied on the Windows Snipping Tool for taking screenshots and doing basic annotation. It's actually not that bad, and recent Windows updates have made Snipping Tool even better: it now includes Quick Markup, a color picker, and screen recording. However, it still falls short compared to my current favorite, Greenshot, a lightweight, free, and open-source screenshot tool. Greenshot lets me do more than just basic editing Obfuscation, auto-cropping, speech bubbles, it's all there One of the most glaring Snipping Tool misses is the lack of a good image editor. Yes, you can click the Edit in Paint option and make your edits there, but it's an added step to something that should be there from the get-go. What you do get in the Snipping Tool itself is basic markup, cropping, and optical image recognition (OCR). Granted, these features are usually enough for the average user who seldom takes screenshots, but they don't nearly cut it for users who take screenshots frequently and want to modify them in more ways. Greenshot comes with a more fleshed-out editor, which is impressive considering how lightweight the tool is. One of my favorite features is the ability to obfuscate sensitive information in the editor itself: this lets me hide my email, the IMEI, and other sensitive device info when taking readers through certain Windows Settings that display them. Additionally, there's a counter that lets me number multiple steps in one screenshot. And this is just the tip of the iceberg. Greenshot also lets you add speech bubbles, text boxes, shapes (or draw freehand), and effect presets to your screenshots. Again, you may not use these features all the time, but
Apr 5, 2026 · via makeuseof.com
Angela Lipps was at her Tennessee home babysitting in July when armed federal agents showed up and arrested her in connection with a string of bank fraud incidents in North Dakota.
Just a moment.
We are getting your experience ready.
Angela Lipps was at her Tennessee home babysitting in July when armed federal agents showed up and arrested her in connection with a string of bank fraud incidents in North Dakota.
Just a moment.
We are getting your experience ready.
Apr 5, 2026 · via nbcnews.com
Abstract As a key indicator of structural integrity and in-service performance, crack detection is essential for the condition assessment and preventive maintenance of bridges. To address the challenges of detecting cracks with various scales and shapes under low-contrast backgrounds in bridge inspection tasks, this paper proposes a robust detection method named CLGDS. It is based on YOLO11 with enhanced feature fusion and SIoU loss optimization, which effectively improves the accuracy and robustness of crack identification. The proposed framework includes three key innovations. (1) A Cross Stage Partially Large Separable Kernel Attention (C2LSKA) module is integrated in the backbone network to enhanced the representation of crack features in the case of morphologically diverse and complex background interference. (2) A Gathering and Distributing (GD) mechanism serves as the neck network, facilitating multi-scale feature fusion and improving the detection performance for cracks of varying scales and geometrically irregular edges. (3) A Scylla-IoU (SIoU) loss function is introduced to replace the commonly used Complete IoU (CIoU) loss. By explicitly incorporating directional sensitivity and multi-scale adaptability, SIoU effectively mitigates angle-dependent misalignment during bounding box regression. Experimental results demonstrate that CLGDS achieves a mean average precision (mAP@50) of 93.5%, outperforming YOLOv5, YOLOv8, and YOLO11 by margins of +1.5%, +0.6%, and +1.2%, respectively. Furthermore, it attains a mAP@50-95 of 68.5%, significantly higher than that of YOLOv5 (61.3%), YOLOv8 (62.6%), and YOLO11 (65.3%). These results validate the effectiveness of CLGDS in accurate bridge crack detection, providing a solid technical foundation for automated structural health monitoring and preventive maintenance. Similar content being viewed by others Data availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. References Zinno, R., Haghshenas, S. S., Guido, G. & VItale, A. Artificial intelligence and structural health monitoring of bridges: A review of the state-of-the-art. IEEE
Apr 5, 2026 · via nature.com
Hong Kong International Airport (HKIA) has switched on 12 additional Face Easy e-Channels, bringing the arrival concourse total to 26 and sharply expanding the city’s fully touch-less immigration capacity. Unlike traditional e-Channels that require an ID-card swipe and fingerprint check, the upgraded lanes authenticate travellers solely through facial recognition. Once the gate opens, clearance takes under seven seconds—around a third of the legacy process. The Immigration Department says the roll-out will continue in phases until 2027, when 52 biometric lanes will replace all ageing kiosks. For organisations coordinating frequent trips through HKIA, specialist services like VisaHQ can streamline the visa side of the journey. Their online portal (https://www.visahq.com/hong-kong/) walks applicants through Hong Kong employment, dependant and business visa requirements, freeing travel managers to focus on flight bookings and biometric enrolment logistics. With passenger throughput at HKIA rebounding to 61 million in 2025, automation is critical to keeping corporate travellers moving: peak-hour queues dropped by an average of 18 minutes after the first batch of lanes went live last September. Employers with frequent-flyer staff should ensure workers update their smart ID cards and consent to facial-image use; holders of valid dependants’ and employment visas are automatically eligible. The technology will be integrated with One-ID boarding later this year, allowing end-to-end curb-to-gate processing without paper documents. Privacy concerns remain subdued—Hong Kong’s Personal Data Privacy Commissioner reports no complaints linked to the system so far—but companies may wish to update internal travel policies to reflect biometric opt-out options for sensitive assignments. For organisations coordinating frequent trips through HKIA, specialist services like VisaHQ can streamline the visa side of the journey. Their online portal (https://www.visahq.com/hong-kong/) walks applicants through Hong Kong employment, dependant and business visa requirements, freeing travel managers to focus on flight bookings and biometric enrolment logistics. With passenger throughput at HKIA rebounding
Apr 5, 2026 · via visahq.com
Abstract Conventional end-to-end deep neural networks often degrade under domain shifts and require costly retraining when deployed in unpredictable, noisy environments. Inspired by biological brains, we propose a modular framework where each module is a recurrent neural network pretrained via a simple, task-agnostic protocol to learn robust, transferable features. This shapes stable yet flexible low-dimensional representations as invariant input-driven continuous attractor manifolds embedded in high-dimensional latent space across different tasks, supporting robust transfer and resilience to temporal perturbations. At deployment, only a lightweight adapter needs training, allowing rapid adaptation to new tasks. Validated on gesture and rehabilitation action recognition tasks, our framework achieves accuracy competitive with state-of-the-art methods, especially in few-shot settings, while requiring an order of magnitude fewer parameters and minimal training. By integrating biologically inspired attractor dynamics with cortical-like modular composition, the framework offers a practical path toward robust, continual adaptation in real-world information processing. Data availability All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. The customized RGB rehabilitation action dataset76 is available at https://doi.org/10.5281/zenodo.16454040 and https://doi.org/10.5281/zenodo.16473362. Code availability Computer code for all simulations and analysis of the resulting data is available at https://doi.org/10.5281/zenodo.16441066. References Xing, W., Li, M., Li, M. & Han, M. Towards robust and secure embodied AI: a survey on vulnerabilities and attacks. Preprint at arXiv https://doi.org/10.48550/arXiv.2502.13175 (2025). Liu, J. et al. Towards out-of-distribution generalization: a survey. Preprint at arXiv https://doi.org/10.48550/arXiv.2108.13624 (2021). Roy, N. et al. From machine learning to robotics: challenges and opportunities for embodied intelligence. Preprint at arXiv https://doi.org/10.48550/arXiv.2110.15245 (2021). Jaeger, H. & Haas, H. Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 304, 78–80 (2004). Barry, C., Hayman, R., Burgess, N. & Jeffery, K. J. Experience-dependent rescaling of entorhinal grids. Nat. Neurosci. 10, 682–684 (2007). Anderson, M.
Apr 4, 2026 · via doi.org
04 Apr MACHINE LEARNING EXPERT WITNESS TESTIMONY FOR ML & AI TRIAL TESTIFYING NEEDS Machine learning expert witnesses for trial testifying note that ML is a subset of AI focused on systems that improve through experience. In legal contexts, reviewers and thought leaders analyze algorithms, assess predictive models, and of course also as best machine learning expert witnesses evaluate the integrity of data used in ML systems. Thought leaders’ expertise helps courts understand whether models function as intended, comply with standards, or caused errors or harm. Top machine learning expert witnesses tend to have advanced knowledge in data science, statistics, and computer programming. SMEs and KOLs interpret training data, model accuracy, overfitting, and algorithmic biases. Famous machine learning expert witnesses get retained in patent disputes, software contract cases, product liability claims, and regulatory investigations involving predictive analytics, recommendation systems, autonomous systems, or data-driven decision-making tools. Via providing independent analysis, detailed reports, and testimony in court, any given machine learning expert witness helps clarify whether models are reliable, ethical, and accurate. Folks turn technical jargon into accessible explanations and can cross-examine opposing experts to ensure impartial evaluation of ML claims. 50 Sample Products and Areas Machine Learning Expert Witnesses Cover - Credit scoring systems - Fraud detection models - Autonomous vehicles - Predictive maintenance systems - Speech recognition software - Image recognition models - Chatbots - Recommendation engines - Algorithmic trading platforms - Smart home automation - Robotics - Personalized marketing tools - Facial recognition applications - AI-driven HR tools - Autonomous drones - Self-driving forklifts - Predictive healthcare software - AI for legal research - Sentiment analysis tools - E-commerce analytics - Customer churn prediction - AI supply chain systems - Predictive policing tools - Insurance claim models - Stock market forecasting algorithms - Natural language processing systems - AI
Apr 4, 2026 · via futuristsspeakers.com
Residents rallied outside Troy City Hall on Friday over the city's contract with Flock and its license-plate reading cameras.
Mayor Carmella Mantello issued a public safety emergency on Wednesday to fund the company's contract, after the city council ordered the auditor to pause payments.
Mantello accuses the council of inappropriate interference with public safety.
City Council President Sue Steele argues Mantello is violating the city charter.
Protesters raised concerns over citizen privacy.
“Flock also does human detection. Flock also does sound detection. These are all getting fed into the same national database. And where’s the line with public safety and privacy? I think that we have a lot of questions about that,” Dierdre Shea of Troy said.
Flock says it only looks at vehicles and does not use facial recognition.
Apr 4, 2026 · via spectrumlocalnews.com
This Week’s Awesome Tech Stories From Around the Web (Through April 4) Every week, we scour the web for important, insightful, and fascinating stories in science and technology. Image Credit NASA/Joel Kowsky Share Artificial Intelligence How AI Helped One Man (and His Brother) Build a $1.8 Billion CompanyErin Griffith | The New York Times ($) "From his house in Los Angeles, Mr. Gallagher, 41, used AI to write the code for the software that powers his company, produce the website copy, generate the images and videos for ads and handle customer service. ...This year, they are on track to do $1.8 billion in sales." Computing The First Quantum Computer to Break Encryption Is Now Shockingly CloseKarmela Padavic-Callaghan | New Scientist ($) "A quantum computer capable of breaking the encryption that secures the internet now seems to be just around the corner. Stunning revelations from two research teams outline how it could happen, with one suggesting that the current largest quantum machine is already more than halfway towards the size needed." Space Four Astronauts Are Now Inexorably Bound for the MoonEric Berger | Ars Technica "For NASA and the Artemis II crew members, [Thursday's main engine burn] marked a point of no return for more than a week. About three-quarters of the American population has not witnessed humans leaving low-Earth orbit in their lifetimes. The last time this occurred was 1972, with the final Apollo Moon mission." Be Part of the Future Sign up to receive top stories about groundbreaking technologies and visionary thinkers from SingularityHub. Computing New Fiber-Optic Record Allows 50,000,000 Movies to Be Streamed at OnceMatthew Sparkes | New Scientist ($) "Faster speeds have been achieved before in highly regulated experiments, but this work crucially used existing cables that have been heavily used, have dirty connectors, sit underneath a
Apr 4, 2026 · via singularityhub.com
Technology - Home - Technology - News OkCupid settles claims it shared user photos with a facial recognition company Dating app OkCupid agreed to settle claims from the Federal Trade Commission that it deceived millions of users by sharing their photos with a third-party facial recognition company without their consent. OkCupid and parent company Match Group did not admit w… Published a day ago on Apr 4th 2026, 5:00 am By Web Desk Dating app OkCupid agreed to settle claims from the Federal Trade Commission that it deceived millions of users by sharing their photos with a third-party facial recognition company without their consent. OkCupid and parent company Match Group did not admit wrongdoing as part of the settlement, but instead promised not to make similar alleged misrepresentations in the future. According to the FTC complaint, after facial recognition company Clarifai reached out to one of OkCupid’s founders in 2014, the app gave it access to nearly three million OkCupid user photos, alongside demographic and location data about users. That access violated OkCupid’s own privacy policy, the FTC alleged, since it didn’t give users a chance to opt out of their data being shared. After sharing the data, OkCupid and Match later tried to obscure their relationship with Clarifai when The New York Times reached out about it for a story, the FTC alleged. Still, the settlement does not impose penalties on OkCupid or Match, nor directly address the data allegedly shared with Clarifai. The companies promise not to misrepresent their data collection and sharing policies in the future, and submit to compliance monitoring, which could subject them to further action if they’re found to violate the order, once approved by a court. FTC consumer protection bureau director Christopher Mufarrige said in a statement that the settlement shows, “The FTC
Apr 4, 2026 · via gnnhd.tv
Abstract Agricultural pests continue to impose serious threats to global food security by causing major yield losses across diverse cropping systems, making early and accurate identification vital for effective pest management. With the growing integration of digital technologies in modern agriculture, deep-learning–based pest recognition has become a promising approach to surpass the limitations of manual scouting and conventional monitoring practices. This work presents a comprehensive experimental evaluation of multiple deep-learning architectures for multi-class pest classification via image-level classification covering 19 pest species. The study investigates classical CNNs (MobileNetV2, VGG16), compound-scaled models (EfficientNetB0/B3, EfficientNetV2-B0), residual architectures (ResNet50), automated NAS models (Xception, NASNetLarge), and novel hybrid CNN–Transformer designs including Hybrid InceptionResNetV2, Hybrid ResNet50 + CBAM, Hybrid EffNet-Transformer, and Hybrid EfficientNetV2-S + Transformer. To enhance foreground isolation and reduce background complexity in field images, segmentation-driven preprocessing is employed using GrabCut, Watershed, SLIC, and Felzenszwalb, generating structure-refined image representations for downstream classification. Results show that attention-augmented hybrid models consistently outperform standalone CNNs, with the Hybrid EfficientNetV2-S + Transformer achieving the highest performance with 0.8800 validation accuracy, 0.849 macro-F1, and 0.4560 validation loss. These findings highlight the effectiveness of combining convolutional feature hierarchies with global self-attention for reliable multi-species pest classification and offer meaningful guidance for developing intelligent precision agriculture systems. Similar content being viewed by others Data availability The datasets analysed during the current study are available in the repository [https://www.kaggle.com/datasets/ibrahimagabardiop/pestaidatasetv2]. References About | Plant Production and Protection | Food and Agriculture Organization of the United Nations. Plant-Production-and-Protection. (2025). https://www.fao.org/plant-production-protection Savary, S., Ficke, A., Aubertot, J. N. & Hollier, C. The global burden of pathogens and pests on major food crops. Nat. Ecol. Evol. 3(3), 430–439. https://doi.org/10.1038/s41559-018-0793-y (2019). Junaid, M. D. & Gokce, A. F. Global agricultural losses and their causes. Bull. Biol. Allied Sci. Res. 2024(1), 66–66 (2024). Ullah, Q. et al. Innovative
Apr 4, 2026 · via nature.com
Two black crows in a German lab have just shaken up what we thought we knew about animal intelligence. In a study published in 2025, carrion crows learned to spot a single odd shape hidden among five nearly identical ones. They did this even when the shapes were tricky quadrilaterals that humans normally meet in math class. For a long time, researchers suspected that this kind of geometric intuition belonged only to our species. So what exactly did the birds do? Perched in front of a touchscreen, each crow saw six shapes at a time. Five were the same. One was the intruder. When the crow pecked the odd one out, a feeder dropped a reward such as a mealworm. At first the differences were obvious, like a crescent among stars. Then the team at the University of Tübingen made the task much tougher with warped squares, skewed diamonds and other quadrilaterals that differed only in subtle angles and side lengths. Even on their very first encounters with these new quadrilateral sets, both crows chose the intruder far above chance. In a game where random pecks would be right only about one time in six, the birds landed on the correct answer close to half the time, and sometimes more. They also did better with very regular shapes such as neat squares and worse when everything was irregular. That pattern looks a lot like the way human volunteers respond in similar tests. For animal cognition researchers, there is an extra twist. Earlier work with baboons using related stimuli suggested that nonhuman primates did not show the same sensitivity to geometric regularity. Some scientists took that as a hint that intuitive shape geometry might be a human specialty. The new crow results suggest that bird brains, which lack a mammalian cortex,
Apr 4, 2026 · via ecoticias.com
Abstract In the field of security X ray imaging, existing datasets mainly focus on endogenous domain shifts caused by hardware differences but do not address pseudocoloring driven domain shift (PDS), a type of domain shift resulting from different pseudocoloring schemes across devices. To fill this gap, we construct and release a benchmark dataset designed for studying domain shifts in security X ray inspection, focusing on both PDS and endogenous differences. The dataset has two complementary subsets: a synthetic subset, which applies controlled hue transformations on an endogenous domain shift dataset to isolate and evaluate PDS independently; and an aggregated subset, which integrates multiple public security datasets into a unified format for verifying cross domain and cross device model generalization. All annotations follow standardized structures with detailed descriptions to ensure reproducibility and comparability. This dataset provides a benchmark for evaluating object detection algorithms under domain shifts and supports studies on model robustness and generalization in diverse X ray security scenarios. Similar content being viewed by others Data availability The dataset is available at Figshare29. The PDSXray dataset and analysis code are publicly available at https://doi.org/10.6084/m9.figshare.29958461. The dataset includes X-ray security inspection data for studying the impact of pseudocoloring-driven (PDS) and endogenous (EDS) domain shifts on model performance. The code contains training and analysis scripts, along with model weights, supporting result reproduction. Both the dataset and code are distributed under the Creative Commons Attribution 4.0 International License (CC BY 4.0) (license details: https://creativecommons.org/licenses/by/4.0/). Code availability The code used in this study is hosted on Figshare alongside the PDSXray dataset29: https://doi.org/10.6084/m9.figshare.29958461. It includes scripts for analyzing PDS, cross-dataset category mapping, and model training, along with model weights for replication. The code is licensed under the CC BY 4.0 license, the same as the dataset. The link is public and will remain available after
Apr 4, 2026 · via nature.com
CNA938 Rewind - Ditch the inkpads? The rise of facial recognition at checkpoints #SGToday
Singapore will roll out facial recognition clearance for motorcyclists at Woodlands Checkpoint from end-March, with Tuas to follow - but could your face be your new thumbprint? Daniel Martin finds out more from Professor Alex Kot, Emeritus Professor, NTU.
Resume
Pause
16 min
Apr 4, 2026 · via channelnewsasia.com
Aiarty’s Image Matting V2.7 and Video Enhancer V3.6 enhance RAW support, AI background removal, and video rendering for faster, more reliable workflows. CHENGDU, SICHUAN, CHINA, March 30, 2026 /EINPresswire.com/ — Aiarty, a leading provider of AI-powered creative solutions, today announced the latest upgrades for Aiarty Image Matting V2.7 and Aiarty Video Enhancer V3.6. These new releases focus on professional AI image and video enhancement, delivering better RAW photo compatibility, improved AI background removal accuracy, and stable AI video rendering for long-duration projects. These updates support photographers, videographers, and digital creators, enhancing AI video enhancement, AI image matting, and batch RAW photo processing for more efficient workflows. In conjunction with these new releases, Aiarty is launching a limited-time Easter promotion, offering up to 49% off lifetime plans for its AI image and video tools: https://www.aiarty.com/store.htm?ttref=w4bd-aiverele-zhh2603-enpr Aiarty Image Matting V2.7: Optimized RAW Background Removal for Photos from Professional Cameras Aiarty Image Matting is an AI background remover that enables photographers to quickly and precisely remove image backgrounds while preserving fine details and textures. Version V2.7 introduces optimized RAW parsing, enhancing compatibility with major RAW formats. It significantly improves color reproduction for Canon CR3 files, widely used by photographers with Canon EOS R series and EOS 5D Mark IV cameras, and refines display logic for RW2 files from popular Panasonic Lumix models like the GH5, S5, and G9, preserving details and tones more faithfully. These improvements make AI background removal and batch RAW photo processing faster, more precise, and more reliable, helping professional photographers achieve high-quality results with minimal manual adjustments. Additional fixes in V2.7 include : * Windows NVIDIA GPU compatibility: Resolves TensorRT inference engine issues after driver updates, ensuring stable AI-powered image processing. * TIFF preview bug fix: Eliminates low-quality thumbnails and improves efficiency for batch RAW workflows. These updates make
Apr 4, 2026 · via blufftontoday.com
Alcatraz closes $50M Series B, plans new verticals and international market expansion California-based biometric access control firm Alcatraz AI has closed a US$50 million Series B funding round, which it will use to expand into new verticals and international markets. The latest round brings the startup’s total funding to over $100M since its founding. The financing was led by Bulgarian private equity firm BlackPeak Capital, Warsaw-based venture fund Cogito Capital, and Taiwania Capital, a venture capital firm founded by the Taiwanese government. Other backers include existing investors Almaz Capital, the European Bank for Reconstruction and Development (EBRD) and Ray Stata, an investor and co-founder of Analog Devices. The company’s main product is the Rock, a face biometric access control system that authenticates users as they walk into a space. The system is anonymized, meaning it doesn’t identify a person based on a stored facial image; instead, it converts snapshots of faces into encrypted representations, which are then bound to a credential. In 2025, the company reported strong growth, including a 200 percent increase in new enterprise customers and a 300 percent year-over-year increase in data center adoption. Other clients include sport venues, R&D centers, financial institutions, universities and airports, including the Honolulu Daniel K. Inouye International Airport. “The world’s largest airports, energy companies, and the world’s most critical data centers all trust Alcatraz,” the company’s CEO, Tina D’Agostin, says in a statement. The fresh financing comes just as the Cupertino-headquartered firm celebrates its 10th anniversary. Alcatraz was founded in 2016 by Vince Gaydarzhiev, a former Apple product lead who worked on hardware prototyping for iPad and iPhone during the development of Face ID. The Bulgarian‑born engineer says he was inspired to “create a Face ID for the physical world,” he recently told media outlet Entrepreneur. The company completed its $25
Apr 4, 2026 · via biometricupdate.com