Computer Vision news and analysis on Web Pulse: 3357 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
Researchers from MIT, NYU, and Sakana AI replicated the Picbreeder experiment using vision-language model agents, finding that while AI can generate diverse images without predefined goals, it struggles to match human cr…
Meta plans to add a facial recognition feature called NameTag to its smart glasses, allowing wearers to identify people and access information via the Meta AI assistant, according to internal documents and code found in …
NVIDIA GPUs deliver eight standout graphics and AI features including Ray Tracing, Path Tracing, DLSS, ShadowPlay, and RTX HDR, according to a July 9, 2026 BGR roundup. The features differentiate NVIDIA's consumer and cr…
Mistral unveiled Robostral Navigate, an 8-billion-parameter model for robot navigation using a single RGB camera and natural language instructions. The model achieves a 76.6% success rate on the R2R-CE benchmark, outperf…
Meta has prototyped smart glasses that continuously record audio and take photos every few seconds, with plans to disable the LED indicator light that currently signals recording. The company's "super sensing" mode would…
FableCut, a browser-based non-linear video editor, exposes its entire timeline as a JSON document, allowing AI agents to edit video via MCP/REST while humans work simultaneously. The zero-dependency tool features Premier…
Hugging Face's top AI papers for July 9, 2026, highlight four key directions: world models for robotics, unified multimodal systems, ultra-long context, and faster inference at deployment. Notable papers include RynnWorl…
Meta has prototyped "super sensing" smart glasses that continuously capture audio and photos every few seconds, enabling AI-powered recall of the wearer's day, according to the Financial Times. The company is considering…
A series of images labeled 'AI or Not' challenge viewers to distinguish between AI-generated and real photographs, focusing on landscapes and street photos.
Mill, a startup that raised over $232 million, is developing an AI-powered commercial bin that dehydrates and analyzes food waste from restaurants and grocery stores. The device reduces waste volume by 80% and uses compu…
Enabled Intelligence, a Virginia-based data-labeling company, added over 500,000 hours of Ukraine conflict drone full-motion video to its EView library, pre-labeled for AI training. The dataset includes electro-optical, …
A new interactive game called 'AI or Not: Sports' challenges users to distinguish between AI-generated and real sports images. The game features photorealistic architecture images and prompts users to guess whether each …
A developer created Mobilerun, an open-source framework that uses LLMs with vision to automate mobile apps that lack public APIs. The tool reads screen content via accessibility trees and vision models, plans multi-step …
Meta is testing a 'super sensing' mode for its Ray-Ban smart glasses that could enable continuous environmental recording, raising privacy concerns. The company sold over 7 million pairs in 2025, signaling growing consum…
A new online game called 'AI or Not' challenges users to distinguish between AI-generated and real photographs across categories like macro, landscape, and street photography, testing the limits of photorealistic AI imag…
LingBot-World 2.0 (LingBot-World-Infinity) introduces unbounded interaction horizons, rapid 60fps 720p video generation, diverse interactive elements, and an agentic harness for world modeling. The real-time version is a…
Datalab released Lift, a 9B vision model for schema-first document extraction that directly outputs structured JSON from PDFs and images, claiming 90.2% field accuracy. The tool competes with NuExtract3, LlamaExtract, an…
A developer built NavBot-D1, an open-source quadruped robot platform that integrates mechanical design, Jetson Orin Nano Super, ROS 2, CAN bus, and reinforcement-learning locomotion. The platform is designed for industri…
ByteDance launched Seedream 5.0 Pro on July 8, 2026, positioning it as a direct rival to OpenAI's GPT-Image 2 with features like layer separation, multi-reference fusion, and multilingual text rendering for professional …
A new online game called 'AI or Not' challenges users to distinguish between AI-generated and real photographs across categories such as street photography, portraits, and architecture. The game aims to test and improve …