Computer Vision

Researchers developed MonteRET, a region-aware retrieval-enhanced framework that improves automated chest CT report generation by integrating global CT features with region-level anatomical representations and using a kn…

Researchers propose G2SR, a geometric method for fast and memory-efficient Gaussian-based surface reconstruction from few RGB views, achieving 69-89 reconstructions per second within 203 MB of GPU memory while matching o…

DS@GT ARC achieved third place in the PlantCLEF 2026 challenge with a multi-scale ViT inference pipeline that blends a fine-tuned DINOv2 ViT-L/14 classifier with FAISS kNN retrieval and habitat-fit priors, reaching a pri…

A new automated red-teaming framework using a multi-agent architecture reduced the False Negative Rate (FNR) from 41.2% to 24.5% on a public image safety benchmark without human labeling, according to a paper on arXiv (2…

A new dynamic manipulation hypergraph framework for vision-based human activity recognition improves hand-object interaction F1 scores by 6.9 percentage points on EPIC-KITCHENS-100/VISOR and 9.5 points on Assembly101 ove…

Researchers introduce Hy-Embodied-RxBrain, an embodied cognition foundation model that jointly performs language-visual reasoning and imagination for planning. The model uses a unified multimodal Mixture-of-Transformers …

A new evaluation framework called Just Keep Prompting (JKP) reveals that repeatedly challenging vision-language models (VLMs) with follow-up questions destabilizes their answers rather than improving reasoning, according…

Researchers propose ReportMedSAM, a report-driven segmentation framework that uses a learnable concept bank and frozen medical vision-language encoder (BiomedCLIP) to align organ-level embeddings with clinical corpora vi…

A developer discovered that AI QA agents using Chrome MCP can produce false positives because browsers throttle requestAnimationFrame and timers in hidden tabs, causing dynamic features like canvas rendering to appear fu…

Blur & Unblur AI is a browser-based tool that detects faces in photos and applies blur only to selected face masks, running entirely locally without uploading images to a server. The tool supports JPG, PNG, and WebP file…

LingBot-Map, a feed-forward 3D foundation model for streaming scene reconstruction, achieves state-of-the-art performance at approximately 20 FPS on 518×378 resolution over sequences exceeding 10,000 frames. The model's …

Kansas has established an AI working group within the Governor's office, currently in the study phase with policy recommendations pending. Recent developments include senators probing workforce development partnerships a…

Hundreds of US cities that adopted Flock Safety's AI license plate detection cameras are now canceling contracts amid privacy concerns and data-sharing with ICE, according to experts and the ACLU. A University of Washing…

Researchers introduce Visual Concept Inference from Sets (VICIS), a task that evaluates vision-language models' ability to infer shared concepts from example images and apply them to new inputs. State-of-the-art VLMs per…

Former DeepMind researcher Andrew Dai raised $55 million in seed funding for his startup Elorian AI at a $300 million valuation before building any product, with backing from Menlo Ventures, Altimeter, and Striker Ventur…

Roboflow released a multi-model auto-labeling workflow for segmentation that combines SAM 3, Google Gemini, and OpenAI's GPT to produce pixel-perfect masks via a 2-of-3 consensus rule, eliminating uncorrelated errors and…

Wyoming gun activists are increasingly alarmed over AI-powered surveillance cameras that search for guns, arguing the technology violates Second Amendment and privacy rights. Companies like Volt and ZeroEyes provide gun-…

RoboFlow's serverless inference API handles thousands of customer-trained models on a shared GPU fleet by solving three core challenges: ingress inversion (multi-megabyte images ingested and reduced to kilobytes of detec…

Timeline Scan, a new service from an unnamed developer, automatically corrects the dates on scanned family photos by analyzing visual clues, handwritten notes, and file metadata, allowing users to organize their librarie…

A developer detailed how to build high-speed vision analyzers on Android by orchestrating data movement between the camera sensor, RAM, and hardware accelerators like NPUs, GPUs, and DSPs to minimize latency-to-insight. …