ExFuse: Enhancing Feature Fusion for Semantic Segmentation

Abstract

Modern semantic segmentation frameworks usually combine low-level and high-level features from pre-trained backbone convolutional models to boost performance. In this paper, we first point out that a simple fusion of low-level and high-level features could be less effective because of the gap in semantic levels and spatial resolution. We find that introducing semantic information into low-level features and high-resolution details into high-level features is more effective for the later fusion. Based on this observation, we propose a new framework, named ExFuse, to bridge the gap between low-level and high-level features thus significantly improve the segmentation quality by 4.0\% in total. Furthermore, we evaluate our approach on the challenging PASCAL VOC 2012 segmentation benchmark and achieve 87.9\% mean IoU, which outperforms the previous state-of-the-art results.

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Rethinking Atrous Convolution for Semantic Image Segmentation
• Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
• Multi-Scale Context Aggregation by Dilated Convolutions
• Feature Pyramid Networks for Object Detection
• Aggregated Residual Transformations for Deep Neural Networks
• Pyramid Scene Parsing Network
• BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation
• Checkerboard artifact free sub-pixel convolution: A note on sub-pixel convolution, resize convolution and convolution resize
• Stacked Deconvolutional Network for Semantic Segmentation
• Understanding Convolution for Semantic Segmentation
• Convolutional Neural Pyramid for Image Processing
• Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes

Cited by in corpus (17)

• Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
• FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation
• Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation
• Knowledge Adaptation for Efficient Semantic Segmentation
• Boundary-Aware Network for Fast and High-Accuracy Portrait Segmentation
• AUNet: Attention-guided dense-upsampling networks for breast mass segmentation in whole mammograms
• Understanding and Improving Encoder Layer Fusion in Sequence-to-Sequence Learning
• DenseAttentionSeg: Segment Hands from Interacted Objects Using Depth Input
• SegSort: Segmentation by Discriminative Sorting of Segments
• Feature Fusion Encoder Decoder Network For Automatic Liver Lesion Segmentation
• Collaborative Boundary-aware Context Encoding Networks for Error Map Prediction
• A^2Net: Adjacent Aggregation Networks for Image Raindrop Removal
• EfficientHRNet: Efficient Scaling for Lightweight High-Resolution Multi-Person Pose Estimation
• Multi Receptive Field Network for Semantic Segmentation
• ESFNet: Efficient Network for Building Extraction from High-Resolution Aerial Images
• SparseMask: Differentiable Connectivity Learning for Dense Image Prediction
• Residual Pyramid FCN for Robust Follicle Segmentation