Computer Vision news and analysis on Web Pulse: 1843 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
Chinese electric vehicle maker Xpeng is spending approximately $41 million per month, or nearly $500 million annually, on AI training for its autonomous driving systems. The company's head of autonomous driving, Dr. Xian…
Artist Trevor Paglen has published a new book, *How to See Like a Machine: Images After AI*, arguing that machine learning has fundamentally altered the function of images, creating an "indexical flip" where any photogra…
Arm announced several Arm-based innovations in May 2026, including new profiling tools for AI and cloud workloads through Arm Performix, which helps developers optimize workloads on Arm Neoverse platforms like Google Clo…
Pinterest committed $4 billion to Amazon Web Services for cloud services through 2031, the largest infrastructure deal in the company's history, to power AI-driven visual search and personalization for its 600 million mo…
Python MSS, a pure Python module using ctypes, now offers ultra fast cross-platform multiple screenshots with no dependencies and thread-safe operation for Python 3.9+. The module captures screenshots per monitor or all …
Apple is testing camera-equipped AirPods that would allow Siri to "see" the wearer's surroundings for navigation, shopping, and AI-driven tasks, according to Bloomberg. The earbuds, which are in late-stage testing with A…
Google published research demonstrating that a phone can track a user's heart rate by visually analyzing their face, offering a simpler alternative to smartwatch-based monitors. The work highlights the potential for phon…
A developer built Spot UI, a local-first browser extension and server that allows AI coding agents to read and act on live, precise UI input. The tool bridges the gap between what developers see on screen and what AI age…
Intel is developing a new AI technology called "frame extrapolation" that predicts a player's next move to generate frames before they are rendered, Intel fellow Tom "TAP" Petersen confirmed in an interview. The feature,…
Google shipped Nano Banana Pro (formerly Gemini 3 Pro Image) to general availability in June 2026, positioning it as the most capable reasoning-driven image model with a public API. The model, priced at $0.134 per 1K or …
South Korea will require all internet communities and forum owners to scan every user-uploaded image and video using AI censorship tools, starting July 1st, under revised telecommunications laws. Website operators must p…
Nouri launched an AI-powered nutrition and fitness platform that scans food, generates personalized meal plans, and adjusts workouts based on users' dietary intake. The service, accessible as a progressive web app on iPh…
Researchers have developed a deep learning algorithm that restores three-dimensional capillary anatomy from a single optical coherence tomographic angiography (OCTA) volume, addressing imaging artifacts that hinder relia…
Researchers developed a sensor-guided adaptive contrastive learning framework that uses privileged LiDAR data during training to teach visual encoders to focus on navigation-relevant geometry rather than scene-specific a…
Researchers developed a phase-aware large language model agent that optimizes retrieval system parameters by conditioning each proposal on its full optimization history, overcoming the limitations of traditional hyperpar…
Researchers developed and evaluated deep learning models for automated age-related macular degeneration (AMD) staging using OCT and OCT angiography data from 271 participants. A biomarker-based model achieved the highest…
Researchers released Biomazon, a 20-meter multimodal benchmark dataset covering the Amazon Basin that pairs GEDI relative height (RH) profiles and aboveground biomass density (AGBD) targets with satellite and land cover …
Researchers have developed RePHO, a method that reconstructs physically plausible human-object interactions from monocular videos by combining kinematic estimates with reinforcement learning. The approach addresses commo…
Researchers have developed NIV (Neural Axis Variations), a method that automatically converts a static font into a fully functional variable font by predicting per-point displacements along desired design axes. Trained o…
Researchers introduced Flash-WAM, a modality-aware distillation framework that compresses world-action model inference to a single step, reducing per-chunk latency from 8.1 seconds to 348 milliseconds on NVIDIA L40S hard…