Direction-aware Spatial Context Features for Shadow Detection

Abstract

Shadow detection is a fundamental and challenging task, since it requires an understanding of global image semantics and there are various backgrounds around shadows. This paper presents a novel network for shadow detection by analyzing image context in a direction-aware manner. To achieve this, we first formulate the direction-aware attention mechanism in a spatial recurrent neural network (RNN) by introducing attention weights when aggregating spatial context features in the RNN. By learning these weights through training, we can recover direction-aware spatial context (DSC) for detecting shadows. This design is developed into the DSC module and embedded in a CNN to learn DSC features at different levels. Moreover, a weighted cross entropy loss is designed to make the training more effective. We employ two common shadow detection benchmark datasets and perform various experiments to evaluate our network. Experimental results show that our network outperforms state-of-the-art methods and achieves 97% accuracy and 38% reduction on balance error rate.

Accepted for oral presentation in CVPR 2018. The journal version of this paper is arXiv:1805.04635

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• Amulet: Aggregating Multi-level Convolutional Features for Salient Object Detection
• Shadow Optimization from Structured Deep Edge Detection
• Fast Shadow Detection from a Single Image Using a Patched Convolutional Neural Network

Cited by in corpus (25)

• Direction-aware Spatial Context Features for Shadow Detection and Removal
• Mask-ShadowGAN: Learning to Remove Shadows from Unpaired Data
• Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications
• Revisiting Shadow Detection: A New Benchmark Dataset for Complex World
• SAC-Net: Spatial Attenuation Context for Salient Object Detection
• Physics-based Shadow Image Decomposition for Shadow Removal
• Progressive Glass Segmentation
• Camouflaged Object Segmentation with Distraction Mining
• Unsupervised Portrait Shadow Removal via Generative Priors
• Instance Shadow Detection with A Single-Stage Detector
• Large-Field Contextual Feature Learning for Glass Detection
• Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset
• Optimizing the F-measure for Threshold-free Salient Object Detection
• MBA-RainGAN: Multi-branch Attention Generative Adversarial Network for Mixture of Rain Removal from Single Images
• Language-Driven Interactive Shadow Detection
• Instance Shadow Detection
• Unsupervised Shadow Removal Using Target Consistency Generative Adversarial Network
• Deep Relighting Networks for Image Light Source Manipulation
• From Shadow Segmentation to Shadow Removal
• Unveiling Deep Shadows: A Survey and Benchmark on Image and Video Shadow Detection, Removal, and Generation in the Deep Learning Era
• Temporal Feature Warping for Video Shadow Detection
• Rain Removal and Illumination Enhancement Done in One Go
• Defocus Blur Detection via Depth Distillation
• Region Refinement Network for Salient Object Detection
• A+D Net: Training a Shadow Detector with Adversarial Shadow Attenuation