ParseNet: Looking Wider to See Better

TL;DR

ParseNet introduces a simple global context augmentation to deep networks for semantic segmentation, significantly improving accuracy with minimal additional computation, and achieves state-of-the-art results on multiple benchmarks.

Contribution

The paper proposes ParseNet, a method that incorporates global average features into convolutional networks, enhancing segmentation performance beyond existing baselines.

Findings

Achieves state-of-the-art on SiftFlow and PASCAL-Context.

Improves baseline performance significantly with global features.

Maintains low additional computational cost.

Abstract

We present a technique for adding global context to deep convolutional networks for semantic segmentation. The approach is simple, using the average feature for a layer to augment the features at each location. In addition, we study several idiosyncrasies of training, significantly increasing the performance of baseline networks (e.g. from FCN). When we add our proposed global feature, and a technique for learning normalization parameters, accuracy increases consistently even over our improved versions of the baselines. Our proposed approach, ParseNet, achieves state-of-the-art performance on SiftFlow and PASCAL-Context with small additional computational cost over baselines, and near current state-of-the-art performance on PASCAL VOC 2012 semantic segmentation with a simple approach. Code is available at https://github.com/weiliu89/caffe/tree/fcn .