MARS: Mask Attention Refinement with Sequential Quadtree Nodes for Car Damage Instance Segmentation
This work addresses the need for more accurate car damage segmentation in the insurance industry, representing an incremental advancement over existing methods.
The paper tackles the problem of insufficient accuracy in car damage instance segmentation for insurance applications by introducing MARS, which uses self-attention with sequential quadtree nodes to refine masks, achieving improvements of +1.3 maskAP with R50-FPN and +2.3 maskAP with R101-FPN on a Thai car-damage dataset compared to SOTA methods.
Evaluating car damages from misfortune is critical to the car insurance industry. However, the accuracy is still insufficient for real-world applications since the deep learning network is not designed for car damage images as inputs, and its segmented masks are still very coarse. This paper presents MARS (Mask Attention Refinement with Sequential quadtree nodes) for car damage instance segmentation. Our MARS represents self-attention mechanisms to draw global dependencies between the sequential quadtree nodes layer and quadtree transformer to recalibrate channel weights and predict highly accurate instance masks. Our extensive experiments demonstrate that MARS outperforms state-of-the-art (SOTA) instance segmentation methods on three popular benchmarks such as Mask R-CNN [9], PointRend [13], and Mask Transfiner [12], by a large margin of +1.3 maskAP-based R50-FPN backbone and +2.3 maskAP-based R101-FPN backbone on Thai car-damage dataset. Our demos are available at https://github.com/kaopanboonyuen/MARS.