Computer Vision news and analysis on Web Pulse: 4319 curated articles tracking the latest Computer Vision developments, tools, and research, updated continuously from vetted sources.
A regression framework built on open-source Vision-Language Models reached a Spearman correlation of 0.67 with a four-model configuration when scoring Olympic diving videos from the AQA-7 benchmark, according to an arXiv…
A decomposed confidence layer combining perception, layout, and validation channels with conformal risk control raised AUROC for financial-document key-value extraction from 0.54-0.74 with native Vision Language Model ve…
Researchers introduced SOLAR, a multimodal generative model that analyzes tomato leaf disease across six question-answering tasks by aligning visual features with task-aware language representations through a mixture-of-…
A new arXiv paper (2609.19384v1) proposes Riemannian–Lorentz Parameter Fusion (RLPF), a geometry-aware method for merging a Vision Transformer and a state-space model into a two-branch hybrid that reaches 82.37% on CIFAR…
A deep-learning perception framework using the YOLOv12-m-seg model achieved a segmentation AP@0.5 of 83.7% with an inference time of 20.4 ms per image, the most favorable balance of accuracy and speed among direct segmen…
A Retrieval-Augmented Generation framework for scientific image quality assessment took 1st place in the SIQA-U understanding track of the SIQA challenge at the ICME 2026 Grand Challenges, according to arXiv paper 2609.1…
A prospective graduate student is weighing two unnamed grad school options that would steer research toward either general large language model work in interpretability, inference, and alignment or agentic and physical A…
Midjourney posted a Senior Machine Learning Engineer role on its clinical team in the San Francisco Bay Area, a hybrid position focused on owning tissue-class segmentation and labeling models for an ultrasound CT clinica…
Scaleout Systems, a Swedish startup founded in 2018 by Uppsala University researchers and selected for NATO's DIANA Challenge Program in 2025, has pivoted from commercial-vehicle machine learning to defense applications,…
Roboflow reported that OpenAI's GPT-6 Astra, released in early September 2026, ranks #1 on its Vision Evals overall and on object detection as of September 17, 2026, and can output per-instance polygons for segmentation …
A developer outlined an architecture for building a hands-free vision AI assistant that pipes camera frames from Meta AI glasses through a Flutter companion app to a secure backend running a Gemini vision model, then ret…
Skylark Labs announced a technology partnership with Skydio to connect video streams from Skydio drones to Kepler, Skylark Labs' self-learning AI analytics platform, for critical infrastructure customers. Skydio already …
Vidu S2 launched with two modes, S2-Avatar for real-time digital humans and S2-Editing for live video edits, both running at 720p. The model uses a Backbone-Refiner architecture plus SRF (Self-Replay Forcing) and a VLM A…
A developer launched S-Roll, a macOS app that performs AI video understanding, transcription, and clipping entirely on-device, according to the Show HN post. S-Roll lets users chat with long-form video to locate specific…
A synthetic data augmentation pipeline built on Amazon SageMaker AI and Amazon Rekognition improved person-detection mAP50 by up to 160 percent for industrial safety models without manual annotation or hazardous photogra…
A two-person team built SENTINEL, a single-file, fully offline on-device posture attention map that runs YOLO26n at roughly 16 FPS at 720p on an M4 Mac with WiFi physically disabled. Engineer Igor Eduardo and field speci…
Researchers are pursuing in-context robot learning with vision-language model (VLM) agents to let robots adapt to unfamiliar environments, according to a report on the work. The approach aims to address the limitation th…
A developer reports that the specialized vision-driven browser agent Mano-CUA 1.1 scored 41.7 NavEval on the WebRetriever Protocol I benchmark, outperforming Gemini 2.5 Pro Computer Use at 40.9 and Claude 4.5 Computer Us…
CaptionQA, a caption-based memory approach that segments egocentric video into 30-to-60-second windows, outperformed direct VideoQA on 10 of 12 models for videos longer than 20 minutes, according to an arXiv paper (2609.…
A developer building an MRI/CT interpretation tool found that left/right laterality errors were the most common failure across four vision models (Claude, Gemini, Grok, and Google's MedGemma) tested on eight real studies…