Channel charting (CC) is a self-supervised learning technique which aims to construct a lower-dimensional representation of the channel measurements, while preserving the neighboring relationship of users. In this letter, we propose a machine learning approach for radio-based localization task with the aid of CC, which fully exploits the dissimilarity extracted solely from the channel measurements. To this end, a hybrid model structure inspired by physical principles is employed, which is realized by a computationally efficient two-layer neural network initialized with a channel chart. The training process employs a weakly-supervised approach that combines: 1) a Siamese network architecture preserving relative user neighborhood relationship through channel dissimilarity metrics, and 2) a limited set of anchor points with ground-truth location annotations to establish absolute positional references. The proposed approach is empirically validated on realistic channel data, achieving encouraging localization accuracy compared to benchmark approaches.
Tianxiang Xu, Li You, Jue Wang et al.· IEEE Wireless Communications...· 0 citations
Multi-view multi-label learning typically suffers from dual data incompleteness due to limitations in feature storage and annotation costs. The interplay of heterogeneous features, numerous labels, and missing information significantly degrades model performance. To tackle the complex yet highly practical challenges, we propose a Theory-Driven Label-Specific Representation (TDLSR) framework. Through constructing the view-specific sample topology and prototype association graph, we develop the proximity-aware imputation mechanism, while deriving class representatives that capture the label correlation semantics. To obtain semantically distinct view representations, we introduce principles of information shift, interaction and orthogonality, which promotes the disentanglement of representation information, and mitigates message distortion and redundancy. Besides, label-semantic-guided feature learning is employed to identify the discriminative shared and specific representations and refine the label preference across views. Moreover, we theoretically investigate the characteristics of representation learning and the generalization performance. Finally, extensive experiments on public datasets and real-world applications validate the effectiveness of TDLSR.
Quanjiang Li, Tianxiang Xu, Tingjin Luo et al.· Neural Information Processin...· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
