Computer Vision news and analysis on Web Pulse: 3325 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
A developer's comparison of seven AI-powered OCR tools in 2026 evaluates them across three use cases: web tools, APIs, and self-hosted solutions. The analysis highlights Google Vision API, Mistral OCR, and ABBYY FineRead…
Banana Pi launched the BPI-AI2N system-on-module and carrier board on March 19, 2025, pairing Renesas' RZ/V2N Vision AI MPU with dual camera inputs, dual Gigabit Ethernet and expandable storage. The company rates the acc…
Researchers at arXiv introduced a novel pre-training method for logographic character recognition that uses semantic-based contrastive learning to improve visual representations on imbalanced datasets. The method, which …
Researchers propose a multimodal machine-learning framework for calibration monitoring and quality assessment in archaeological digitisation, integrating photogrammetric 3D reconstruction, hyperspectral imaging, X-ray fl…
A new arXiv study (2608.00077v1) audits visual-token pruning in OCR-critical multimodal large language models (MLLMs), finding that accuracy alone misses failures where correct answers lack local token provenance. On loc…
A new study from arXiv (arXiv:2608.00135v1) evaluating Convolutional Neural Networks on the JONES-19 cultural design dataset finds that learning from scratch with repeated local sampling (multi-crop) can match the discri…
Researchers introduced SPARC-Rad, a manually curated multimodal benchmark dataset and evaluation pipeline for assessing spatial and anatomical reasoning in radiology vision-language models (VLMs). The dataset includes 30…
Researchers from an unnamed institution tested hybrid quantum convolutional neural networks (QCNNs) with 2 and 4 qubits to classify volcanic clouds in SEVIRI satellite images, comparing their performance to classical arc…
Researchers introduced ELECTRIC (Evidential Learning-Enhanced CT Reconstruction via Iterative Correction), a physics-guided Bayesian method that uses an evidential neural network to iteratively reconstruct CT images. In …
Researchers propose LeapTalk, a framework for real-time talking-head generation that achieves high-fidelity video synthesis in a single forward step at up to 200 FPS, addressing the latency-quality trade-off in existing …
Researchers propose DAR, a reference-guided renderer that extends Wan2.2 camera control to a joint camera-plus-geometry interface, enabling pretrained video diffusion models to act as 4D renderers from animated meshes. O…
Researchers propose a ControlNet-based diffusion framework for map-to-satellite image synthesis that conditions on OpenStreetMap raster maps and their stationary wavelet transform (SWT) subbands, using two adapters fused…
A developer has published a bypass for the Ghost Font protection mechanism on Mixfont, using a non-machine-learning attack that achieves a 0.9077 IoU mask overlap and reads 'GHOST'. The same attack against a hardened ver…
A developer has created a workflow combining Claude Code with a Chrome extension to automatically identify and fix visual discrepancies between a UI mockup and the actual rendered screen. The system captures the rendered…
Auterion is equipping 50,000 Ukrainian-made SkyFall Shrike strike drones with its terminal-guidance system under a German-funded order worth about $103 million (90 million euros), with deliveries beginning in mid-July. T…
MarkTechPost published a tutorial on evaluating multimodal vision models using Moonshot PerceptionBench, a benchmark that measures fine-grained visual perception across tasks such as OCR, counting, localization, contextu…
Auterion and Ukrainian manufacturer SkyFall have begun supplying Ukraine with 50,000 Shrike FPV drones equipped with AI-enabled Skynode S systems, a contract worth about 90 million euros ($103 million) funded by Germany,…
Procore will acquire DroneDeploy for approximately $845 million in cash, a deal announced on July 29th and expected to close in 2026, giving the construction software provider control of drones, robots, and cameras that …
Google has identified a firmware bug in Walmart's Onn smart cameras that disables familiar face detection for Google Home Premium subscribers, and a fix is being validated for release in the coming weeks. The issue, repo…
Researchers present a training algorithm that reduces optimization variance in small neural networks used for real-time neural shading, improving convergence reliability. The method uses an ensemble of network instances,…