InspectVLM: Unified in Theory, Unreliable in Practice
This work addresses the reliability of VLMs for industrial inspection, highlighting practical deployment challenges, and is incremental in evaluating an existing paradigm.
The paper tackled the problem of using unified vision-language models (VLMs) for industrial inspection tasks, finding that while InspectVLM performed competitively on some tasks like classification and keypoint localization, it failed to match traditional models in core inspection metrics, exhibiting brittle behavior and degenerate outputs.
Unified vision-language models (VLMs) promise to streamline computer vision pipelines by reframing multiple visual tasks such as classification, detection, and keypoint localization within a single language-driven interface. This architecture is particularly appealing in industrial inspection, where managing disjoint task-specific models introduces complexity, inefficiency, and maintenance overhead. In this paper, we critically evaluate the viability of this unified paradigm using InspectVLM, a Florence-2-based VLM trained on InspectMM, our new large-scale multimodal, multitask inspection dataset. While InspectVLM performs competitively on image-level classification and structured keypoint tasks, we find that it fails to match traditional ResNet-based models in core inspection metrics. Notably, the model exhibits brittle behavior under low prompt variability, produces degenerate outputs for fine-grained object detection, and frequently defaults to memorized language responses regardless of visual input. Our findings suggest that while language-driven unification offers conceptual elegance, current VLMs lack the visual grounding and robustness necessary for deployment in precision critical industrial inspections.