Computer Vision news and analysis on Web Pulse: 4315 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
A developer has built SignBridge, a prototype that generates videos of individual words being signed in American Sign Language from text prompts. The developer, who has years of ASL experience in the Deaf community, repo…
A developer built a browser-based skin lesion screening app that runs entirely on-device using WebGPU acceleration, Transformers.js, and WebLLM, so no images are uploaded to a server. The app loads a quantized Vision Tra…
University of Florida researchers found that simple repeated visual patterns, such as black-and-white stripes or a checkerboard-like projection, can cause stereo-camera depth-estimation systems in autonomous vehicles, dr…
An AI traffic camera in Bengaluru issued a Rs 500 helmet fine to a motorcycle rider after mistaking a guitar for a pillion rider, according to an e-challan received by the rider traveling from Hebbal towards Tin Factory.…
BlueVetaUpright, a purpose-built product-photo orientation system, reached 99.02% accuracy on clean cardinal rotations, 96.67% across the full 360° range, and 84.65% on rotations exactly ±40° from any clean angle, accord…
OpenAI's GPT-6 Astra can identify incorrectly assembled IKEA furniture from a photo with 80 percent accuracy, up from 28 percent for the best model in November 2025, according to Epoch AI. Epoch AI noted the model's spee…
A tutorial demonstrates an end-to-end multimodal data augmentation and adversarial robustness workflow using Meta's AugLy library for images, text, and audio, connected directly to PyTorch datasets and DataLoaders. The w…
A developer extended the Jev-style single-function LLM wrapper to support vision models by adding an `attachments` field for base64 images to the request format, then built a webcam demo that classifies frames with plain…
OpenRouter's public model catalog listed 180 of 625 routes, or 29 percent, producing non-text output such as images, video, speech, transcripts and embeddings as of September 25, 2026, according to a snapshot taken by th…
Liquid AI released LFM2.5-VL-3B-DSpark, an experimental speculative-decoding draft model for its LFM2.5-VL-3B vision-language model that adds about 279.5M parameters and delivers up to 3.13x faster decoding on Apple sili…
A developer built an open-source orienteering analysis portal in which an AI agent wrote all of the code, with the human only setting tasks and validating results on real competitions. The portal georeferences scanned co…
A developer debugging an Android barcode scanner found that a CameraX preview stayed visually correct after a tablet was rotated 180°, but the frames delivered to the ImageAnalysis analyzer were upside down. After haltin…
Reactor and the Amazon Neuron Science team developed a kernel-centric optimization strategy that enables real-time, low-latency video generation on AWS Trainium chips, according to Reactor co-founder and CTO Bryce Schmid…
Tancky AI launched AI ImageCombiner, a web tool that merges 2 to 4 uploaded JPG, PNG, or WebP images (up to 16 MB each) into a single photorealistic scene using generative neural vision. The tool accepts an optional text…
RushShift launched an offline AI desktop tool for Windows that indexes local video and photo archives and lets editors search them by natural-language prompt, matching on visual meaning rather than filenames. The tool ru…
A developer built Handrail, an open-source (Apache-2.0) screen assistant that splits visual question-answering and UI element localization into two separate vision model calls against the same screenshot, finding that co…
Alibaba's DAMO Academy has open-sourced DAMO RADAR, a vision-language AI model that identifies 146 clinical findings, including several cancers, from contrast-enhanced abdominal CT scans covering 18 organs. In a study pu…
A developer released a reusable agent skill called creative-breakdown that detects whether an input is a photo, website screenshot, ad, or video and routes it to the open-source vision model Ling-3.0-flash-VL on OpenRout…
Vincent Sitzmann and co-authors submitted an arXiv paper on 17 June 2020 proposing sinusoidal representation networks, or Sirens, which use periodic activation functions for implicit neural representations. The authors r…
Omneky published Taste Bench, a blind evaluation of 8 leading image models across 59 real ad briefs spanning 4 brands and 5 languages, scored through 1,371 blind reviews. OpenAI's GPT Image 2.5 Flare and GPT Image 2.5 Su…