mlsanityData to 7 Oct 2026

A Hybrid Vision-Language Architecture for Automated Defect Reasoning and Report Generation in Industrial Inspection

arXiv 2605.26533 ↗First posted 26 May 2026cs.CV cs.AI cs.CLcited by 0

On the map: LLM reasoning → Chain-of-thought reasoning for driving

Abstract (arXiv)

Automated industrial inspection requires both precise defect localization and structured maintenance report generation; in current practice these tasks are handled separately, with linguistic interpretation left to human experts. This paper describes a decoupled, edge-deployable pipeline for wind turbine blade inspection built from three components that each handle a distinct sub-task. The Eyes a YOLO26-x-obb oriented bounding-box detector localizes defects at dataset-native resolution. The Bridge a deterministic, parameter-free encoding module maps each detected bounding box to grid-referenced spatial tokens embedded in a structured prompt. The Brain a 4-bit quantized Qwen-2.5-1.5B model adapted with Quantized Low-Rank Adaptation (QLoRA) on 947 synthetically generated maintenance reports generates a structured JSON report from that prompt. Retrieval-Augmented Fine-Tuning (RAFT) further grounds each recommendation in indexed maintenance procedures. Five ablation experiments, scored by BLEU-4, ROUGE-L, Hallucination Rate (HR), and an LLM-as-a-Judge rubric, compare the pipeline against a monolithic vision-language model (VLM) baseline and against partial configurations in which one component is removed. The complete system achieves BLEU-4 0.41, HR=4%, and Expert Score = 8.6/10 compared with 0.07, 65%, and 3.3/10 for the zero-shot VLM baseline. The QLoRA-adapted 1.5B model generates higher-quality reports than a 671B-parameter generalist API model given identical detection evidence, at 47 tokens per second on a single T4-class GPU. The results show that purpose-built decoupled architecture with a small domain-specific training corpus outperforms a generalist end-to-end model on this structured generation task.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Chain-of-thought reasoning for driving, one of 17 directions in LLM reasoning. Builds on 13 of its 45 references that are in the map. First posted on arXiv 26 May 2026.

Builds on

13 of 45 references.

References · 13

  1. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  2. 2023Vision-Language Models for Vision Tasks: A Surveycited by 834
  3. 2023AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Modelscited by 169
  4. 2024A Review on Edge Large Language Models: Design, Execution, and Applicationscited by 98
  5. 2025YOLO26: Key Architectural Enhancements and Performance Benchmarking for Real-Time Object Detectioncited by 76
  6. 2024Effectiveness Assessment of Recent Large Vision-Language Modelscited by 29
  7. 2024FabGPT: An Efficient Large Multimodal Model for Complex Wafer Defect Knowledge Queriescited by 12
  8. 2023Image Synthesis under Limited Data: A Survey and Taxonomycited by 9
  9. 2022Identification of Surface Defects on Solar PV Panels and Wind Turbine Blades using Attention based Deep Learning Modelcited by 6
  10. 2026MAU-GPT: Enhancing Multi-type Industrial Anomaly Understanding via Anomaly-aware and Generalist Experts Adaptationcited by 6
  11. 2025LLM-Guided Evolution: An Autonomous Model Optimization for Object Detectioncited by 5
  12. 2025A domain adaptation neural network for digital twin-supported fault diagnosiscited by 1
1 more
  1. 2025Cognitive-YOLO: LLM-Driven Architecture Synthesis from First Principles of Data for Object Detectioncited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Seeing the Unseen: Towards Training-Free Inspection for Wind Turbine Blades Using Knowledge-Augmented Vision Language Models
  2. 2026GenAU: Language-Grounded Industrial Anomaly Understanding with Vision-Language Models
  3. 2025Towards Open-Vocabulary Industrial Defect Understanding with a Large-Scale Multimodal Dataset
  4. 2026Unification of Closed-Open Industrial Detection Scenarios: New Large-Scale Benchmarks,Challenges and Baselines
  5. 2026Synthetic Defect Image Generation for Power Line Insulator Inspection Using Multimodal Large Language Models
  6. 2025InspectVLM: Unified in Theory, Unreliable in Practice
  7. 2026Industrial Anomaly Detection via Defect-Grounded Reasoning in Visual Latent Space
  8. 2026SteelDefectX: A Multi-Form Vision-Language Dataset and Benchmark for Steel Surface Defect Analysis