A Graph Neural Network-Based Dual Attention Fusion Network for CSI-Based Activity Recognition
Bibliographic record
Abstract
Over the past decade, Channel State Information (CSI)-based human activity recognition (HAR) has attracted wide attention. Despite significant advancements, existing CSI-based HAR methods primarily face two critical challenges: 1) how to exploit intrinsic hierarchical spatial correlations spanning adjacent sub-carriers while maintaining global awareness of entire CSI series; 2) how to establish a cross-dimensional (e.g., spatial, temporal) optimization framework that enables effective information fusion across distinct feature domains to achieve robust CSI series prediction. To address these challenges, we propose Wi-DualAtt, a novel graph neural network(GNN)-based CSI feature extraction network, specifically designed for the effective fusion of spatial and temporal dimensions. The proposed Wi-DualAtt is composed of three key components: a graph attention network (GAT)-based hierarchical correlation attention network (GHCAN), a temporal feature attention network (TFAN), and a prediction fusion module (PFM). Specifically, GHCAN employs spatial attention to capture the hierarchical correlation among all sub-carriers. Meanwhile, TFAN utilizes an attention layer to extract significant temporal features from CSI samples. Finally, PFM integrates the recognition results from the aforementioned two components, utilizing a knowledge distillation mechanism to form the final recognition result, thereby enhancing the recognition capability for CSI-based HAR systems. Extensive experimental results demonstrate that Wi-DualAtt outperforms several state-of-the-art models, achieving recognition accuracy exceeding 99% across various CSI-based activity recognition scenarios.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".