Improving Semantic Image Segmentation with a Probabilistic Superpixel-Based Dense Conditional Random Field

Computer Science

• Image Segmentation 100%
• Conditional Random Field 100%
• Superpixel 100%
• Semantic Image 100%
• Deep Convolutional Neural Networks 75%
• Random Field Model 50%
• Postprocessing 25%
• Segmentation Task 25%
• Segmentation Performance 25%
• Quadratic Programming 25%
• Residual Neural Network 25%
• Computational Efficiency 25%
• Clustering Algorithm 25%
• Object Boundary 25%
• Local Structure 25%
• Visual Attention 25%
• Fully Convolutional Network 25%

Engineering

• Random Field 100%
• Convolutional Neural Network 100%
• Random Field Model 66%
• Optimization Method 33%
• Computational Efficiency 33%
• Object Boundary 33%
• Inference Algorithm 33%
• Pixel Level 33%
• Local Structure 33%
• Clustering Algorithm 33%