Buscar recursos académicos

We propose a technique for producing `visual explanations' for decisions from a large class of Convolutional Neural Network (CNN)-based models, making them more transparent. Our approach - Gradient-weighted Class Activat...

In this work, we revisit the global average pooling layer proposed in [13], and shed light on how it explicitly enables the convolutional neural network (CNN) to have remarkable localization ability despite being trained...

The way that information propagates in neural networks is of great importance. In this paper, we propose Path Aggregation Network (PANet) aiming at boosting information flow in proposal-based instance segmentation framew...