Pixel-Adaptive Convolutional Neural Networks

We propose a pixel-adaptive convolution (PAC) operation, a simple yet effective modification of standard convolutions, in which the filter weights are multiplied with a spatially-varying kernel that depends on learnable, local pixel features. PAC is a generalization of several popular filtering techniques and thus can be used for a wide range of use cases. Specifically, we demonstrate state-of-the-art performance when PAC is used for deep joint image upsampling. PAC also offers an effective alternative to fully-connected CRF (Full-CRF), called PAC-CRF, which performs competitively, while being considerably faster. In addition, we also demonstrate that PAC can be used as a drop-in replacement for convolution layers in pre-trained networks, resulting in consistent performance improvements.

Authors: 
Hang Su (University of Massachusetts, Amherst)
Varun Jampani (NVIDIA)
Deqing Sun (NVIDIA)
Erik-Learned Miller (University of Massachusetts, Amherst)
Publication Date: 
Sunday, June 16, 2019