Metraipolis › archive › dc_55ca88a6f63d

KOALA: Koopman Operator Learning for WiFi-Based Anticipatory Hum — arXiv:2608.15815

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.15815
License (as archived)
CC0-1.0
Archived
2026-10-18T19:40:58.757Z
SHA-256 of the archived text
a53e424bb26d2b4f56120e927730e24ba484005ce41037d4c3011b3973234b35
Stable URL of this page
https://metraipolis.com/documento/dc_55ca88a6f63d

How to cite

KOALA: Koopman Operator Learning for WiFi-Based Anticipatory Hum — arXiv:2608.15815 — arXiv (en), https://arxiv.org/abs/2608.15815 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_55ca88a6f63d (Metraipolis archive dc_55ca88a6f63d, sha256 a53e424bb26d2b4f56120e927730e24ba484005ce41037d4c3011b3973234b35)
BibTeX
@misc{dc55ca88a6f63d,
  title = {KOALA: Koopman Operator Learning for WiFi-Based Anticipatory Hum — arXiv:2608.15815},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.15815},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_55ca88a6f63d; sha256 a53e424bb26d2b4f56120e927730e24ba484005ce41037d4c3011b3973234b35},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.15815. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

WiFi Channel State Information (CSI) has emerged as a privacy-preserving alternative to cameras for human pose estimation. However, existing approaches treat pose inference as an instantaneous regression problem and do not model temporal dynamics, making future motion prediction infeasible. Naively applying vision-based prediction methods compounds the estimation noise already present in CSI-derived poses, as autoregressive rollouts amplify errors at every step. We propose KOALA, the framework for human motion prediction directly from WiFi CSI, by lifting noisy CSI-derived pose sequences into a learned Koopman latent space where nonlinear dynamics become linear, enabling multi-horizon prediction via simple matrix-vector products without autoregressive iteration or error accumulation. A residual CSI-conditioned operator resolves the identity attractor problem inherent from Koopman formulations, and an anchor-delta prediction head eliminates the degenerate shortcut of copying the current pose across all horizons. To regularise the lifting and operator jointly, we introduce a Koopman Anchored Latent (KAL) loss that operates in the temporal-encoder feature space, enforcing dynamical consistency across prediction horizons without requiring contrastive, spectral, or auxiliary losses. Experiments on MM-Fi and WiPose show that KOALA achieves robust, consistent performance across both short- and long-term prediction horizons, outperforming all baselines by a substantial margin.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.15815 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T19:02:44.125Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).