RF-ray: Joint RF and Linguistics Domain Learning for Object Recognition
summary
System uses an Impinj reader, one antenna and 7×7 tag array of 49 UHF passive tags, collecting RSS and phase across 7 frequencies and 6 power settings; a sensing‑enhancement module, channel attention, multi‑scale perceptron, environment discriminator, U‑Net for shape and CNN+FC for material produce joint shape and material outputs. Prototype achieves >99% overall, unseen shape‑material pairs >92% accuracy and >97% mIOU for shape, and 92.9% accuracy for zero‑shot linguistic embedding based unseen material prediction.
AI-written summary · about
related talks
10mRadiological AGI: How Close Are We?
17mLabelAR: A Spatial Guidance Interface for Fast Computer Vision Image Collection
23mVision-Language Models for Fraud Detection
19mDocument Reader using Deep Learning with a Small Dataset
01h 00mOpening Keynote
28mVisual Search
20mInteracting with Soli: Exploring Fine-Grained Dynamic Gesture Recognition in the Radio-Frequency ...
30mStarting deep learning projects without enough labelled data
