SLICK: Selective Localization and Instance Calibration for Knowledge-Enhanced Car Damage Segmentation in Automotive Insurance
This work addresses the need for accurate damage detection in real-world automotive inspection workflows, though it appears incremental as it builds on existing segmentation methods with domain-specific enhancements.
The paper tackles the problem of precise car damage segmentation for automotive insurance by introducing SLICK, a framework that leverages structural priors and domain knowledge, achieving superior segmentation performance and robustness in experiments on large-scale datasets.
We present SLICK, a novel framework for precise and robust car damage segmentation that leverages structural priors and domain knowledge to tackle real-world automotive inspection challenges. SLICK introduces five key components: (1) Selective Part Segmentation using a high-resolution semantic backbone guided by structural priors to achieve surgical accuracy in segmenting vehicle parts even under occlusion, deformation, or paint loss; (2) Localization-Aware Attention blocks that dynamically focus on damaged regions, enhancing fine-grained damage detection in cluttered and complex street scenes; (3) an Instance-Sensitive Refinement head that leverages panoptic cues and shape priors to disentangle overlapping or adjacent parts, enabling precise boundary alignment; (4) Cross-Channel Calibration through multi-scale channel attention that amplifies subtle damage signals such as scratches and dents while suppressing noise like reflections and decals; and (5) a Knowledge Fusion Module that integrates synthetic crash data, part geometry, and real-world insurance datasets to improve generalization and handle rare cases effectively. Experiments on large-scale automotive datasets demonstrate SLICK's superior segmentation performance, robustness, and practical applicability for insurance and automotive inspection workflows.