Computer Vision news and analysis on Web Pulse: 4325 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
Hands-on testing of Google's Gemini video understanding model and the commercial detection API Sightengine found current AI video detection tools cannot reliably distinguish real footage from AI-generated footage, with a…
Hugging Face released Workflow1111, a Gradio Workflow rebuild of AUTOMATIC1111's stable-diffusion-webui feature set as a single canvas graph of eleven media pipelines built from seventy-three nodes. The Space, runnable b…
Roboflow Edge AI Engineer Aadhav Sivakumar demonstrated a missing-part detection system for end-of-line inspection in a webinar, running a custom vision model on an NVIDIA Jetson beside the production line with a Basler …
Researchers introduced RenderFormer-V2, a unified transformer-based neural rendering model designed to complement physics-based rendering systems by handling diverse light-transport effects including caustics, volumetric…
Roboflow Inference with Enterprise Offline Mode lets organizations run computer vision models and Workflows on their own hardware with no public internet access, keeping camera feeds, inference, and production results in…
A developer has released an open-source Python tool, tiktok-subtitle-ripper, that extracts on-screen text overlays and subtitles from TikTok and Reels ads to identify the first three-second hook. The tool uses GPU OCR an…
NVIDIA's Synthetic Video Detector (SVD), part of its AI for Media tools showcased at IBC 2026, is now moving into production, achieving 99.3% accuracy for text-to-video and 97.7% for image-to-video detection. Integration…
At IBC 2026 in Amsterdam, NVIDIA announced an expansion of its NVIDIA AI for Media platform, introducing the NVIDIA Synthetic Video Detector (SVD) NIM microservice with accuracy reaching 99.3% for text-to-video and 97.7%…
A developer detailed the design of iLook, a privacy-first face-analysis API that emphasizes explicit scope, confidence values, and limitations in its responses. The system supports REST, MCP, and A2A protocols while main…
Meta and Panmnesia have jointly proposed a next-generation AI datacenter architecture that operates like a single chip using CXL interconnect technology, as published in Nature Reviews Electrical Engineering. The design …
Elbit Systems, Israel's largest defence contractor, launched Fuse, a new division developing AI-based autonomous military technologies including drones and unmanned systems, operating from Rosh HaAyin, Israel. The unit w…
Researchers propose a method for instruction-guided 3D editing that eliminates the need for paired supervision by distilling generative priors, addressing the scarcity of high-quality training data. The approach aims to …
SKF, a Swedish bearing manufacturer, has implemented a focused AI solution for smart product defect detection, improving product quality in its supply chain. The company's approach demonstrates that targeted AI applicati…
A developer compared eight video-to-animation workflows, including GoEnhance AI, DomoAI, Pollo AI, Runway, Luma Dream Machine, Kaiber, Media.io, and EbSynth, focusing on how each preserves motion, identity, and visual st…
AMD researchers presented a generative method for global illumination at ECCV 2026 that produces temporally stable indirect lighting using a single-step latent diffusion model, outperforming prior methods on PSNR, SSIM, …
Alaya Lab researchers released WorldSculpt, a research system that reconstructs crowded multi-view scenes into hundreds of separate, editable 3D meshes, enabling individual objects to be selected and manipulated independ…
Glance, a new macOS app, brings FaceID-like face unlock to Macs using on-device Apple Vision and Core ML frameworks, converting faces into 512-dimensional embeddings stored locally with AES-GCM encryption. The app, which…
Marigold V2, a diffusion transformer-based model for monocular depth estimation, revisits the approach to improve generalization to out-of-distribution inputs, addressing a persistent challenge in the field with applicat…
ScanChess, a new AI-powered tool, digitizes handwritten chess scoresheets and diagrams into PGN/FEN formats by photographing them, transcribing moves with rule validation, and flagging uncertain entries for review. The t…
Researchers introduced RelightFormer, a feed-forward generative transformer that performs multiview object relighting by integrating multi-view cues to overcome the limitations of traditional inverse rendering and single…