可解释性
稳健性(进化)
计算机科学
清晰
适应性
人工智能
变压器
计算机视觉
模式识别(心理学)
工程类
生态学
生物化学
化学
电压
生物
电气工程
基因
作者
Jieru Mei,Liang-Chieh Chen,Alan Yuille,Cihang Xie
出处
期刊:Cornell University - arXiv
日期:2024-01-01
被引量:3
标识
DOI:10.48550/arxiv.2401.02931
摘要
In this work, we introduce SPFormer, a novel Vision Transformer enhanced by superpixel representation. Addressing the limitations of traditional Vision Transformers' fixed-size, non-adaptive patch partitioning, SPFormer employs superpixels that adapt to the image's content. This approach divides the image into irregular, semantically coherent regions, effectively capturing intricate details and applicable at both initial and intermediate feature levels. SPFormer, trainable end-to-end, exhibits superior performance across various benchmarks. Notably, it exhibits significant improvements on the challenging ImageNet benchmark, achieving a 1.4% increase over DeiT-T and 1.1% over DeiT-S respectively. A standout feature of SPFormer is its inherent explainability. The superpixel structure offers a window into the model's internal processes, providing valuable insights that enhance the model's interpretability. This level of clarity significantly improves SPFormer's robustness, particularly in challenging scenarios such as image rotations and occlusions, demonstrating its adaptability and resilience.
科研通智能强力驱动
Strongly Powered by AbleSci AI