A Transformer is a deep learning architecture based on self-attention mechanisms that enables the model to capture long-range dependencies across sequential or spatial input data without relying on recurrence or convolution. In WiFi CSI sensing, Transformers are applied to learn rich, context-aware representations from CSI time-series and feature sequences, making them particularly effective for tasks such as activity recognition, gesture detection, and human pose estimation where capturing temporal and cross-subcarrier relationships is critical. Key variants relevant to this field include the standard encoder-based Transformer, Vision Transformer (ViT) adapted for CSI spectrograms, and lightweight or compressed Transformer variants designed to reduce computational overhead and support energy-efficient or on-device deployment in green sensing scenarios.
Source Papers
- A Survey on Green Wireless Sensing: Energy-Efficient Sensing via WiFi CSI and Lightweight Learning ↗ — A Survey on Green Wireless Sensing: Energy-Efficient Sensing
- Context-Aware Predictive Coding: A Representation Learning Framework for WiFi Sensing ↗ — Context-Aware Predictive Coding: A Representation Learning F
- Deep Learning-Enhanced Human Sensing with Channel State Information: A Survey ↗ — Deep Learning-Enhanced Human Sensing with Channel State Info
- WiFi-Based Human Sensing With Deep Learning: Recent Advances, Challenges, and Opportunities ↗ — WiFi-Based Human Sensing With Deep Learning: Recent Advances
- WiGNN: WiFi-Based Cross-Domain Gesture Recognition Inspired by Dynamic Topology Structure ↗ — WiGNN: WiFi-Based Cross-Domain Gesture Recognition Inspired