Cross-entropy loss is a standard classification objective function that quantifies the dissimilarity between predicted probability distributions and ground-truth class labels, penalizing confident incorrect predictions more heavily. In WiFi CSI sensing research, it serves as the primary training signal for deep learning models performing discrete recognition tasks such as activity classification, gesture recognition, and crowd counting, enabling models to learn discriminative feature representations from complex channel state information. Variants relevant to this field include standard categorical cross-entropy used in benchmarking frameworks like SenseFi for comparing model architectures, and its integration within few-shot or meta-learning pipelines such as DASECount, where it guides adaptation to new domains with minimal labeled samples.
Source Papers
- DASECount: Domain-Agnostic Sample-Efficient Wireless Indoor Crowd Counting via Few-Shot Learning ↗ — DASECount: Domain-Agnostic Sample-Efficient Wireless Indoor
- RSSI-Assisted CSI-Based Passenger Counting with Multiple Wi-Fi Receivers ↗ — RSSI-Assisted CSI-Based Passenger Counting with Multiple Wi-
- SenseFi: A library and benchmark on deep-learning-empowered WiFi human sensing ↗ — SenseFi: A library and benchmark on deep-learning-empowered
- Towards Environment Independent Device Free Human Activity Recognition ↗ — Towards Environment Independent Device Free Human Activity R
- WiGNN: WiFi-Based Cross-Domain Gesture Recognition Inspired by Dynamic Topology Structure ↗ — WiGNN: WiFi-Based Cross-Domain Gesture Recognition Inspired