Latest news, developments and reporting on Computer Vision.
Inkive, a new iOS app, uses on-device computer vision and OCR to export highlighted paper-book passages to Obsidian as Markdown.
A demo at an art exhibition shows people expect cameras to explain the world. As AI shifts to on-device inference, sensors are becoming the new bottleneck, sparking a race across the entire imaging stack.
Chinese researchers claim a lightweight AI system can identify F-22 and F-35 heat signatures with over 90% accuracy in laboratory tests.
A MarkTechPost tutorial demonstrates an end-to-end document intelligence pipeline with docTR, covering OCR, layout, KIE, and searchable PDF export.
Engadget breaks down photo upscaling methods, from basic interpolation to AI tools, and offers tips on selecting the right image format.