This SHREC’19 track aims to explore a novel and challenging research topic on cross domain 3D object retrieval, which means 2D object-based 3D object retrieval to pair a 2D object in one RGB image captured in real world with the corresponding 3D object designed by CAD software. Theoretically, we would like to bring in the recent popular research topics, e.g. deep learning, cross domain learning, few shot learning, for 3D object retrieval. Moreover, we hope to advocate the application of the related research achievements for emerging areas, such as autonomous driving, automatic checkout system.
Selected participants will be invited to submit their methods to the special issues of several highly reputed international journals, e.g. Signal Processing, Multimedia Tools and Applications.