A multi-encoder fusion-attention wave network based on transformers combines Wi-Fi and 5G data for passive ISAC indoor scene inference, with a real prototype achieving localization errors below 0.6 m in 84% of cases.
WiSigPro: Transformer for elevating CSI-based human activity recognition through attention mechanisms,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FAWN: A MultiEncoder Fusion-Attention Wave Network for Integrated Sensing and Communication Indoor Scene Inference
A multi-encoder fusion-attention wave network based on transformers combines Wi-Fi and 5G data for passive ISAC indoor scene inference, with a real prototype achieving localization errors below 0.6 m in 84% of cases.