Foreground Segmentation Using a Triplet Convolutional Neural Network for Multiscale Feature Encoding

Abstract

A common approach for moving objects segmentation in a scene is to perform a background subtraction. Several methods have been proposed in this domain. However, they lack the ability of handling various difficult scenarios such as illumination changes, background or camera motion, camouflage effect, shadow etc. To address these issues, we propose a robust and flexible encoder-decoder type neural network based approach. We adapt a pre-trained convolutional network, i.e. VGG-16 Net, under a triplet framework in the encoder part to embed an image in multiple scales into the feature space and use a transposed convolutional network in the decoder part to learn a mapping from feature space to image space. We train this network end-to-end by using only a few training samples. Our network takes an RGB image in three different scales and produces a foreground segmentation probability mask for the corresponding image. In order to evaluate our model, we entered the Change Detection 2014 Challenge (changedetection.net) and our method outperformed all the existing state-of-the-art methods by an average F-Measure of 0.9770. Our source code will be made publicly available at https://github.com/lim-anggun/FgSegNet.

This paper is under consideration at Pattern Recognition Letters

References in corpus (5)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• Fully Convolutional Networks for Semantic Segmentation
• A Deep Convolutional Neural Network for Background Subtraction
• Towards Benchmarking Scene Background Initialization

Cited by in corpus (20)

• AUTSL: A Large Scale Multi-modal Turkish Sign Language Dataset and Baseline Methods
• Learning Multi-scale Features for Foreground Segmentation
• Learning to Measure Change: Fully Convolutional Siamese Metric Networks for Scene Change Detection
• Combining Background Subtraction Algorithms with Convolutional Neural Network
• 3DFR: A Swift 3D Feature Reductionist Framework for Scene Independent Change Detection
• BDNet: Bengali Handwritten Numeral Digit Recognition based on Densely connected Convolutional Neural Networks
• Deep Neural Network Concepts for Background Subtraction: A Systematic Review and Comparative Evaluation
• DeepPBM: Deep Probabilistic Background Model Estimation from Video Sequences
• Background Subtraction with Real-time Semantic Segmentation
• Scene Change Detection Using Multiscale Cascade Residual Convolutional Neural Networks
• Deep Learning for Robust Motion Segmentation with Non-Static Cameras
• An Integrated Image Filter for Enhancing Change Detection Results
• Learning to Detect Instantaneous Changes with Retrospective Convolution and Static Sample Synthesis
• Video Segmentation Learning Using Cascade Residual Convolutional Neural Network
• Using a Supervised Method without supervision for foreground segmentation
• Automated Prostate Cancer Diagnosis Based on Gleason Grading Using Convolutional Neural Network
• To See in the Dark: N2DGAN for Background Modeling in Nighttime Scene
• CDN-MEDAL: Two-stage Density and Difference Approximation Framework for Motion Analysis
• MetricUNet: Synergistic Image- and Voxel-Level Learning for Precise CT Prostate Segmentation via Online Sampling
• AmphibianDetector: adaptive computation for moving objects detection