Computer Vision

Black Forest Labs released FLUX 3, a multimodal foundation model that jointly learns from images, video, audio and robot action prediction within a single architecture, generating clips up to 20 seconds with native audio…

Cape Coral, Florida, officials considered equipping sanitation trucks with AI-powered cameras from City Detect to identify code violations such as overgrown lawns and illegal dumping, but the city decided not to move for…

A developer built Brickbeam, an AI app that identifies LEGO pieces from a photo and suggests what can be built with them. The app aims to help users rediscover possibilities in their existing LEGO collections rather than…

Apple plans to unveil its first smart glasses at WWDC 2027 next June, with a focus on privacy features that differentiate them from rivals like Meta Ray-Bans, according to Bloomberg's Mark Gurman. The glasses are expecte…

A security service blocked access to a page about an Edge AI traffic counter using YOLO11 and RK3576, triggered by a potential online attack. The site owner can be contacted for resolution, with the Cloudflare Ray ID pro…

AI safety researchers have formalized Physical Prompt Injection Attacks (PPIA), a new class of exploit where visual payloads presented to a Vision-Language Model's camera lens can hijack an autonomous robot's behavior wi…

A technical analysis of PCVR performance bottlenecks identifies GPU frame timing fluctuations, not VRAM capacity, as the primary cause of micro-stutters that break immersion. The article argues that optimizing AI-driven …

A developer found that Tesseract, a PDF parser first created by HP Labs in 1985, outperforms Anthropic's Claude AI for parsing scanned PDFs, being 10x faster and more accurate. Anthropic's own internal Claude skills use …

AI-powered calorie-tracking apps underestimated calories by 250 to 345 calories per meal on average, according to a study presented at NUTRITION 2026 by researchers from the National Institutes of Health. The four apps t…

Cornell Tech researchers have developed an optical receiver that directly alters a processor's memory using photocurrents from beamed light, enabling AI models to update parameters without power-hungry analog circuits. T…

A developer warns that AI-generated videos with inconsistent physics could alter human perception of reality. The brain's predictive processing model, which learns from real-world interactions, may be disrupted by exposu…

A free tool called Next-Gen Vision Engine generates text prompts from Danbooru pony images using secure serverless backend models. Users can drag and drop an image to extract AI prompts.

Microchip Technology signed a definitive agreement on July 24 to acquire edge AI chipmaker Hailo, though the transaction has not yet closed and financial terms were not disclosed. Microchip expects completion near the en…

Slate, an invoicing and bookkeeping SaaS for Israeli businesses, has integrated five AI features—receipt scanning, a chat assistant, support reply drafter, natural language document creation, and column mapping for Excel…

A team named JONAM won third place at the Kijani Space Hackathon by building an AI-powered web application that uses Copernicus satellite data to monitor water quality in Lake Victoria. The platform analyzes variables su…

Generative AI's threat to human creativity is not unprecedented but mirrors historical reactions to tools like photography, argues a new analysis by Joshua Leigh. The piece traces how moral panic and accusations of 'soul…

A developer building Project AI Vision is implementing a transparent prompt engineering layer that forces image classification models to justify their visual reasoning, aiming to debug hallucinations such as misidentifyi…

MIT has implemented an AI video surveillance system that converts raw video feeds into structured, searchable metadata using large language models. The system captures streams via RTSP or WebRTC, extracts features with Y…

Astrallist launches as a smart assistant for home cooks, meal planners, and creators, featuring AI-powered fridge scanning, real-time pantry management, budget tracking, and 1-click Amazon ordering. The app also includes…

A new technique called 'Decoy Fonts' bypasses Claude's vision system by embedding thin, high-contrast decoy characters that the AI's OCR prioritizes over actual letterforms, causing the model to hallucinate the decoy tex…