Enhancing CNNs robustness to occlusions with bioinspired filters for border completion
Enhancing CNNs robustness to occlusions with bioinspired filters for border completion
We exploit the mathematical modeling of the visual cortex mechanism for border completion to define custom filters for CNNs. We see a consistent improvement in performance, particularly in accuracy, when our modified LeNet 5 is tested with occluded MNIST images.
Catarina P. Coutinho、Aneeqa Merhab、Janko Petkovic、Ferdinando Zanchetta、Rita Fioresi
计算技术、计算机技术
Catarina P. Coutinho,Aneeqa Merhab,Janko Petkovic,Ferdinando Zanchetta,Rita Fioresi.Enhancing CNNs robustness to occlusions with bioinspired filters for border completion[EB/OL].(2025-04-24)[2025-05-25].https://arxiv.org/abs/2504.17619.点此复制
评论