Incorporating Depth into both CNN and CRF for Indoor Semantic Segmentation

2017/05/21 by Jindong Jiang, Zhijun Zhang, Jiang, Jindong +5 · 1 citation
Computer Science · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Video Surveillance and Tracking Methods

paper · pdf · doi:10.48550/arxiv.1705.07383

openalex publication_date 2017/05/21 · openalex created_date 2017/06/05 · openalex updated_date 2026/07/28

Abstract

To improve segmentation performance, a novel neural network architecture (termed DFCN-DCRF) is proposed, which combines an RGB-D fully convolutional neural network (DFCN) with a depth-sensitive fully-connected conditional random field (DCRF). First, a DFCN architecture which fuses depth information into the early layers and applies dilated convolution for later contextual reasoning is designed. Then, a depth-sensitive fully-connected conditional random field (DCRF) is proposed and combined with the previous DFCN to refine the preliminary result. Comparative experiments show that the proposed DFCN-DCRF has the best performance compared with most state-of-the-art methods.

Cited by

• Survey on semantic segmentation using deep learning techniques

Related

• Review on Indoor RGB-D Semantic Segmentation with Deep Convolutional Neural Networks
• Joint Semantic Segmentation and Depth Estimation with Deep Convolutional Networks
• Gaussian Filter in CRF Based Semantic Segmentation
• Depth-Adapted CNNs for RGB-D Semantic Segmentation
• Combining the Best of Convolutional Layers and Recurrent Layers: A Hybrid Network for Semantic Segmentation