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.
Mistral AI released OCR 4.1, a document processing service now in public preview, featuring native paragraph-level bounding box extraction, structural block labels, and block-level confidence scores. The service is price…
Batch Normalization, a technique introduced to stabilize deep neural network training, has become a cornerstone of modern deep learning, particularly in convolutional neural networks for computer vision. The method norma…
Researchers reported a Vision Transformer model on June 11, 2026, that jointly predicts tumor type, TP53 biomarkers, and survival-related outcomes from whole-slide histopathology images across 32 solid cancers. The study…
Siamese neural networks, also known as twin neural networks, use shared weights to compare two input vectors and compute comparable outputs, with applications in face recognition, handwriting recognition, and text matchi…
Mistral AI released OCR 4.1, a public preview of its optical character recognition service for Document AI, featuring native paragraph-level bounding box extraction, structural block labels, and block-level confidence sc…
Liquid AI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model for on-device deployment, averaging 69.4 across 28 vision benchmarks, matching InternVL-3.5-4B and 0.7 points behind Qwen3.5-4B. The model reads scr…
Unconventional AI reported that sparsity in coupled oscillator networks cuts up to 98.4% of connections in its Un-0 image generation model while improving performance by 1.9 FID on ImageNet 64x64. The company detailed th…
A developer built Netra, an edge-AI camera system that autonomously tracks a person using an ESP32-CAM, YOLOv8 object detection, and servo motors coordinated over MQTT. The project features proportional control and a dea…
Sherlock, an AI-powered face search app developed by Popy Akter, has been released on the Google Play Store, allowing users to search for faces using artificial intelligence. The app, available at the provided URL, has g…
A new mobile sign language recognition system combines MediaPipe's 21 hand and 468 face landmarks with a GRU/LSTM sequence classifier, achieving real-time 30fps performance on devices via ONNX or TensorRT optimization. T…
Upscal, a native C++/Vulkan image upscaler for Windows, offers free, offline AI enhancement for images and videos across Windows, macOS, iPhone, and iPad, with no cloud upload, subscription, or hidden costs. The app proc…
An engineer at Ships Itself demonstrated an AI agent that processed 13 invoices and correctly blocked $1,411.25 in fraudulent payments. The system uses a vision model to extract data from invoice images, then relies on d…
A developer built Screenshot Vault, an app that OCRs every screenshot on a phone and makes them searchable. The app uses on-device Google ML Kit for text recognition to keep images private, and sends only extracted text …
An AWS developer integrated Amazon Nova 2 Lite via Amazon Bedrock into an event-driven image processing pipeline to extract text and icon metadata from AWS Builder Cards. The multimodal model is invoked through a simple …
A Wikipedia editor revised the Flock Safety surveillance article to add misuse cases in Milwaukee and Georgia, correct a misquoted student, change a Columbine Valley recording to ring footage, and cite Kansas case number…
Researchers introduced S2R, a closed-loop framework for video reflection removal that includes S2R-Synthesis for physics-grounded reflection simulation, S2R-Removal as the first diffusion-based video reflection removal m…
Researchers introduced Global Style Transfer (GST), a Many-to-One artistic image synthesis paradigm that aggregates multiple artworks from a target artist to transfer their shared global style to a single content image, …
Researchers introduced Semantic Prism, a generative semantic segmentation framework that achieves 72.07% mean intersection over union (mIoU) on the Cityscapes validation set, outperforming direct-interface decoding by 11…
A team led by researchers from multiple institutions won the MeViS-Text track at the 8th Large-scale Video Object Segmentation (LSVOS) Challenge 2026 with their SSUPER pipeline, achieving a Final score of 0.9081339614 on…
Researchers introduced COGENT (Counterfactual Gaussian Explanations), a framework that generates counterfactual explanations in the parameter space of Gaussian-based volumetric representations for medical imaging, built …