Learning to Measure Change: Fully Convolutional Siamese Metric Networks for Scene Change Detection

Abstract

A critical challenge problem of scene change detection is that noisy changes generated by varying illumination, shadows and camera viewpoint make variances of a scene difficult to define and measure since the noisy changes and semantic ones are entangled. Following the intuitive idea of detecting changes by directly comparing dissimilarities between a pair of features, we propose a novel fully Convolutional siamese metric Network(CosimNet) to measure changes by customizing implicit metrics. To learn more discriminative metrics, we utilize contrastive loss to reduce the distance between the unchanged feature pairs and to enlarge the distance between the changed feature pairs. Specifically, to address the issue of large viewpoint differences, we propose Thresholded Contrastive Loss (TCL) with a more tolerant strategy to punish noisy changes. We demonstrate the effectiveness of the proposed approach with experiments on three challenging datasets: CDnet, PCD2015, and VL-CMU-CD. Our approach is robust to lots of challenging conditions, such as illumination changes, large viewpoint difference caused by camera motion and zooming. In addition, we incorporate the distance metric into the segmentation framework and validate the effectiveness through visualization of change maps and feature distribution. The source code is available at https://github.com/gmayday1997/ChangeDet.

arXiv comment10 pages, 12 figures

References in corpus (7)

• Deformable Convolutional Networks
• Holistically-Nested Edge Detection
• Foreground Segmentation Using a Triplet Convolutional Neural Network for Multiscale Feature Encoding
• Semantic Instance Segmentation via Deep Metric Learning
• Learning to Compare Image Patches via Convolutional Neural Networks
• Depth Adaptive Deep Neural Network for Semantic Segmentation
• Learning Dense Convolutional Embeddings for Semantic Segmentation

Cited by in corpus (11)

• DASNet: Dual attentive fully convolutional siamese networks for change detection of high resolution satellite images
• Hierarchical Paired Channel Fusion Network for Street Scene Change Detection
• Guided Anisotropic Diffusion and Iterative Learning for Weakly Supervised Change Detection
• Weakly Supervised Change Detection Using Guided Anisotropic Difusion
• Weakly Supervised Silhouette-based Semantic Scene Change Detection
• Epipolar-Guided Deep Object Matching for Scene Change Detection
• Looking for change? Roll the Dice and demand Attention
• Automatic Map Update Using Dashcam Videos
• DR-TANet: Dynamic Receptive Temporal Attention Network for Street Scene Change Detection
• A Weakly Supervised Convolutional Network for Change Segmentation and Classification
• Contextual Scene Augmentation and Synthesis via GSACNet