FINLOS-TriSeg: Fine-Grained Fisheye Semantic Segmentation for GNSS NLOS Identification in Consumer-Grade Measurement Systems

Keyphrases

• Measurement System 100%
• Global Navigation Satellite System 100%
• Non-line-of-sight (NLoS) 100%
• Semantic Segmentation 100%
• Non-line-of-sight Identification 100%
• Consumer-grade 100%
• Blockage 50%
• Positioning Accuracy 50%
• Boundary Model 50%
• Urban Canyon 50%
• Student Teachers 50%
• Learning Framework 50%
• False Alarm Rate 50%
• Three-class 50%
• Real-time Kinematic (RTK) Positioning 50%
• Vision-based 50%
• Satellite Visibility 50%
• Quality Label 50%
• Foliage 50%
• Segmentation Performance 50%
• Urban Scenario 50%
• Segmentation Framework 50%
• Collaborative Learning 50%
• Challenging Environments 50%
• Reliable Signals 50%
• Complex Urban Environment 50%
• Environmental Perception 50%
• Non-line-of-sight Propagation 50%
• Conventional Signal 50%
• Fisheye Camera 50%
• Environmental Visibility 50%
• Canopy Cover 50%
• Signaling Domain 50%
• Semantic Mask 50%
• Efficiency Trade-offs 50%
• Multi-sensor Fusion Navigation 50%
• Self-supervised Pre-training 50%
• Signal Quality 50%
• Pixel-level Annotation 50%
• Complex Scenes 50%
• Semitransparent 50%
• Lightweight Deployment 50%
• Cross Pseudo Supervision 50%
• Class Perception 50%
• Explicit Boundary 50%

Computer Science

• Satellite System 100%
• Image Segmentation 100%
• Measurement System 100%
• Annotation 50%
• Learning Framework 50%
• Segmentation Performance 50%
• Open Source 50%
• Urban Environment 50%
• Experimental Result 50%
• False Alarm Rate 50%
• Collaborative Learning 50%
• Interface Object 50%

Earth and Planetary Sciences

• GNSS 100%
• Line of Sight 100%
• Environmental Perception 25%
• Real Time 25%
• False Alarm 25%