LayerCAM: Exploring Hierarchical Class Activation Maps for Localization

• Published in: IEEE transactions on image processing Vol. 30; pp. 5875 - 5888
• Main Authors: Jiang, Peng-Tao, Zhang, Chang-Bin, Hou, Qibin, Cheng, Ming-Ming, Wei, Yunchao
• Format: Journal Article
• Language: English
• Published: New York IEEE 2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: class activation maps; Convolution; Feature maps; Image segmentation; Localization; Location awareness; Pixels; Reliability; Semantic segmentation; Semantics; Spatial resolution; Task analysis; Weakly-supervised object localization

The class activation maps are generated from the final convolutional layer of CNN. They can highlight discriminative object regions for the class of interest. These discovered object regions have been widely used for weakly-supervised tasks. However, due to the small spatial resolution of the final...