A Restricted Boltzmann Machine (RBM) is a generative stochastic neural network composed of a visible layer and a hidden layer with symmetric connections but no intra-layer connections, trained to learn a probability distribution over its inputs via contrastive divergence. In the context of WiFi/CSI sensing and crowd analysis, RBMs matter as unsupervised feature extractors capable of capturing latent representations from high-dimensional signal or visual data, serving as building blocks for deeper architectures such as Deep Belief Networks (DBNs) when stacked in sequence. Their ability to model complex, non-linear dependencies in unlabeled data makes them relevant for scenarios where labeled CSI samples are scarce, though they have largely been superseded by end-to-end deep learning approaches that offer stronger temporal and spatial generalization.
Source Papers
- Deep Learning-Enhanced Human Sensing with Channel State Information: A Survey ↗ — Deep Learning-Enhanced Human Sensing with Channel State Info
- Recent trends in crowd analysis: A review ↗ — Recent trends in crowd analysis: A review
- Time matters: Empirical insights into the limits and challenges of temporal generalization in CSI-based Wi-Fi sensing ↗ — Time matters: Empirical insights into the limits and challen