VLMs benchmarked for textile sorting, Qwen leads accuracy

Researchers have developed a digital twin-driven robotic system for automated textile sorting, integrating visual language models (VLMs) for classification and foreign object detection. The system was benchmarked using nine VLMs across various garment types and foreign objects, with the Qwen model family achieving the highest accuracy at 87.9%. Lighter models like Gemma3 were noted for their competitive speed-accuracy trade-offs for edge deployment. This approach combines VLM reasoning with grasp detection and digital twin technology to enhance manipulation reliability and enable scalable, autonomous textile sorting in industrial settings. AI

IMPACT VLMs show promise for industrial automation tasks like textile sorting, with specific models offering different trade-offs for accuracy and deployment.

RANK_REASON The cluster is based on an academic paper detailing a new research methodology and benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• Gemma3
• Qwen
• Visual Language Models

• paper
• product
• other

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