这篇论文提出的FDT模拟人眼注视，只用一半token就比DeiT-S更准还省34%算力，而且天然抗噪声攻击。
该论文提出FDT（Foveated Dynamic Transformer），受人类视觉系统中央凹采样和眼动启发。FDT包含注视模块定位注视点过滤无关信息，中央凹模块生成多尺度中央凹嵌入。在50%固定预算下，FDT在ImageNet上准确率81.9%，高于DeiT-S的80.9%，同时减少34.57%乘加操作。FDT对噪声和对抗攻击展现出天然鲁棒性，无需专门训练。
Foveation-Guided Dynamic Token Selection for Robust and Efficient Vision Transformers
The human visual system (HVS) employs foveated sampling and eye movements to achieve efficient perception, conserving both metabolic energy and computational resources. Drawing inspiration from this robustness and adaptability, we introduce the Foveated Dynamic Transformer (FDT), a foveation-guided dynamic token-selection architecture that integrates these mechanisms into a vision transformer framework. The FDT exhibits strong resilience to various types of noise and adversarial attacks, despite not being explicitly trained for such challenges. This inherent robustness is achieved through the use of fixation and foveation modules: the fixation module identifies fixation points to filter out irrelevant information, while the foveation module generates foveated embeddings with multi-scale information. At the 50% fixation-budget setting, FDT achieves higher accuracy than DeiT-S (81.9% vs. 80.9%) while reducing multiply-accumulate operations by 34.57%, highlighting one operating point on its accuracy-efficiency trade-off. These attributes position FDT as an HVS-inspired step toward artificial neural networks that combine adaptive computation with improved resilience.