A Domain Adaptation Framework for Human Activity Monitoring Using Passive Wi-Fi Sensing
Bibliographic record
Abstract
Due to the omnipresence of radio frequency signals, the Channel State Information (CSI) can offer an alternate source to image, video, and other high-dimensional streams in a great many Internet-of-Things (IoT) applications. As a result, an ever increasing number of researchers are advocating for the use of passive CSI data for ranging, tracking, perception and automation across many domains such as robotics, healthcare, and surveillance. Specifically, in indoor applications where movements cause classifiable effects on the CSI, this resource from the existing infrastructure can be leveraged in order to provide a high-dimensional signal source for activity recognition. However, the task remains a challenge on two accounts. On the one hand, the radio frequency channel is highly susceptible to environment changes and artifacts. On the other hand, there is a lack of robust models that allow for practical deployment. In this work, we focus on tackling these issues for practical IoT applications through a layered architecture proposal. In the first layer, the task of inference for activity recognition is performed by a robust cross-modal deep learning network. In the next layer, a domain adaptation framework is proposed in order to achieve high deployment-time performance in target indoor environments. Our experiments verify that through few-shot training, using the proposed framework the activity classification performance increases substantially in practical settings over the state-of-the-art approaches.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 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".