Pixels, Regions, and Objects: Multiple Enhancement for Salient Object Detection

Keyphrases

• Human Visual System 100%
• Saliency Detection 100%
• Publicly Available 50%
• Pixel Level 50%
• Multi-object 50%
• Feature Enhancement 50%
• Iterative Refinement 50%
• Benchmark Dataset 50%
• Decomposition Mechanism 50%
• Decoder 50%
• Boundary Features 50%
• Comprehensive Evaluation 50%
• Adaptive Features 50%
• Complex Scenes 50%
• Iterative Training 50%
• Region Level 50%
• Global Features 50%
• Salient Object 50%
• Sensibility 50%
• Training Strategy 50%
• Background Clutter 50%
• Detailed Features 50%
• Dual-branch 50%
• Multi-scale Feature Enhancement 50%
• System Mechanisms 50%
• Hybrid Loss 50%
• Content Integrity 50%
• Human Visual Cognition 50%
• Frequency Decomposition 50%
• Object-level 50%
• Feature Sequence 50%
• Cognition Mechanism 50%

Engineering

• Human Visual System 100%
• Level Feature 50%
• Pixel Level 50%
• Multiscale 50%
• Input Feature 50%
• Feature Sequence 50%
• Frequency Decomposition 50%

Computer Science

• Human Visual System 100%
• Salient Object Detection 100%
• Enhancement Module 50%
• Iterative Refinement 50%
• Frequency Decomposition 50%