New LYNRED-MDS dataset targets low-visibility pedestrian detection

Researchers have introduced the LYNRED Mobility Dataset Multimodal Detection Subset (LYNRED-MDS), a new dataset designed to improve early collision prediction in low-visibility driving conditions. This subset of the LYNRED Mobility Dataset contains 4000 RGB-infrared image pairs captured in various weather, lighting, and road scenarios around Grenoble, France. The dataset aims to enhance the generalization capabilities of pedestrian detection systems for advanced driver-assistance systems by covering critical edge cases that existing datasets like FLIR ADAS and LLVIP do not adequately address. AI

IMPACT This dataset could lead to more robust pedestrian detection systems for autonomous vehicles, particularly in challenging weather and lighting conditions.

RANK_REASON New dataset release described in an arXiv paper. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• FLIR ADAS
• Grenoble
• LLVIP
• Loic Arbez
• LYNRED-MDS
• LYNRED Mobility Dataset Multimodal Detection Subset
• YOLOv8n

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →