Metraipolis › archive › dc_5004178a2ff9
Keep the Needle, Prune the Haystack: Defect-Preserving Token Pruning for Efficient Zero-Shot Anomaly Detection — arXiv:2608.03681
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.03681
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T12:59:46.518Z
- SHA-256 of the archived text
6c47a5b199c7c179b143f9d6d07cf5a5dc8ccf491ee212d3a5408a4f70b7fbd4- Stable URL of this page
https://metraipolis.com/documento/dc_5004178a2ff9
How to cite
Keep the Needle, Prune the Haystack: Defect-Preserving Token Pruning for Efficient Zero-Shot Anomaly Detection — arXiv:2608.03681 — arXiv (en), https://arxiv.org/abs/2608.03681 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5004178a2ff9 (Metraipolis archive dc_5004178a2ff9, sha256 6c47a5b199c7c179b143f9d6d07cf5a5dc8ccf491ee212d3a5408a4f70b7fbd4)
BibTeX
@misc{dc5004178a2ff9,
title = {Keep the Needle, Prune the Haystack: Defect-Preserving Token Pruning for Efficient Zero-Shot Anomaly Detection — arXiv:2608.03681},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.03681},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5004178a2ff9; sha256 6c47a5b199c7c179b143f9d6d07cf5a5dc8ccf491ee212d3a5408a4f70b7fbd4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.03681. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Zero-shot visual anomaly detection has achieved remarkable progress, with recent vision-only approaches further improving performance while simplifying the inference pipeline. However, existing methods typically perform dense computation over all images and spatial tokens, despite the fact that normal samples dominate real-world scenarios and anomalies usually occupy only small regions. Token pruning offers a promising solution, but introduces an asymmetric pruning risk in anomaly detection: retaining normal tokens mainly incurs redundant computation, whereas removing anomalous tokens may eliminate the only evidence for detection and localization. This risk is particularly severe in early layers, where pruning provides the greatest computational benefit but anomaly semantics remain unreliable. We propose KeepAD, a defect-preserving token pruning framework that formulates token selection as high-recall, anomaly-aware routing. In shallow layers, KeepAD combines coverage-preserving selection over local $2\times2$ patch neighborhoods with deterministic anomaly rescue to reduce the risk of discarding subtle defects. In deeper layers, frozen normal and abnormal prototypes guide pruning under an image-adaptive token budget, aggressively removing low-risk normal tokens while preserving local anomaly evidence. Dense-to-sparse self-distillation further supervises early token routing without introducing additional inference overhead. Experiments on six industrial and seven medical zero-shot anomaly detection benchmarks show that KeepAD reduces the token retention ratio to below $20\%$, while limiting the average degradation in image-level and pixel-level AUROC to within $2.7$ percentage points. At the most aggressive operating point, KeepAD achieves a $7.9\times$ speedup over the strongest CLIP-based baseline. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-05 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2608.03681 — 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 2026-10-31T12:45:33.269Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-05?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).