3DSNet: Unsupervised Shape-to-Shape 3D Style Transfer
This work addresses the largely unexplored problem of 3D style transfer for researchers and practitioners working with 3D object manipulation and generation, offering a novel learning-based solution.
This paper introduces 3DSNet, the first learning-based approach for 3D style transfer between 3D objects, capable of synthesizing new 3D shapes (point clouds and meshes) by combining content from a source model and style from a target model. The method also learns multimodal style distributions, increasing the variety of styles that can be applied to an input shape.
Transferring the style from one image onto another is a popular and widely studied task in computer vision. Yet, style transfer in the 3D setting remains a largely unexplored problem. To our knowledge, we propose the first learning-based approach for style transfer between 3D objects based on disentangled content and style representations. The proposed method can synthesize new 3D shapes both in the form of point clouds and meshes, combining the content and style of a source and target 3D model to generate a novel shape that resembles in style the target while retaining the source content. Furthermore, we extend our technique to implicitly learn the multimodal style distribution of the chosen domains. By sampling style codes from the learned distributions, we increase the variety of styles that our model can confer to an input shape. Experimental results validate the effectiveness of the proposed 3D style transfer method on a number of benchmarks. The implementation of our framework will be released upon acceptance.