The evolution of anti-visual perception: biomimetics, spectral physics, and artificial intelligence
Abstract The co-evolution of visual perception and anti-perception represents a biological arms race now accelerating as a technological ‘hide-and-seek’ in human society. This review synthesizes anti-visual perception as an emerging multidisciplinary paradigm spanning biological, physical, and computational domains. We examine biomimetic strategies—from morphological camouflage to tunable photonic microarchitectures—transformed into engineered prototypes for precise perception manipulation. We analyze the expansion into multi-spectral domains, elucidating matter-wave interactions and material designs enabling signature management from ultraviolet to microwave. We then explore the algorithmic frontier where anti-perception shifts from physical manipulation to exploiting AI vulnerabilities through adversarial attacks across digital, physical, and multimodal scenarios. Integrating these perspectives, this review charts a coherent trajectory and provides a forward-looking framework to guide next-generation camouflage, stealth, robust sensing, and trustworthy AI—continually redefining the contest between seeing and remaining unseen.
Authors
- Mingyang Li (ORCID: https://orcid.org/0000-0003-3542-7445)
- Li Liu (ORCID: https://orcid.org/0000-0003-4588-4906)
- Panhe Hu
- Haowen Tan
- Dongfang Guan
- Muang Ding
- Xiaolong Su
- Zhen Liu
- Jing Tian
- Yongxiang Liu
Institutions
- National University of Defense Technology (CN)
Publication Details
- Journal
- National Science Review
- Published
- 2026-09-17
- DOI
- https://doi.org/10.1093/nsr/nwag601
- Primary Topic
- Neural Networks and Reservoir Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China