Yuezhi Yang

We introduce a 3D detailizer, a neural model which can instantaneously (in <1s) transform a coarse 3D shape proxy into a high-quality asset with detailed geometry and texture as guided by an input text prompt. Our model is trained using the text prompt, which defines the shape class and characterizes the appearance and fine-grained style of the generated details. The coarse 3D proxy, which can be easily varied and adjusted (e.g., via user editing), provides structure control over the final shape. Importantly, our detailizer is not optimized for a single shape; it is the result of distilling a generative model, so that it can be reused, without retraining, to generate any number of shapes, with varied structures, whose local details all share a consistent style and appearance. Our detailizer training utilizes a pretrained multi-view image diffusion model, with text conditioning, to distill the foundational knowledge therein into our detailizer via Score Distillation Sampling (SDS). To improve SDS and enable our detailizer architecture to learn generalizable features over complex structures, we train our model in two training stages to generate shapes with increasing structural complexity. Through extensive experiments, we show that our method generates shapes of superior quality and details compared to existing text-to-3D models under varied structure control. Our detailizer can refine a coarse shape in less than a second, making it possible to interactively author and adjust 3D shapes. Furthermore, the user-imposed structure control can lead to creative, and hence out-of-distribution, 3D asset generations that are beyond the current capabilities of leading text-to-3D generative models. We demonstrate an interactive 3D modeling workflow our method enables, and its strong generalizability over styles, structures, and object categories.

This paper studies the classical problem of 3D shape alignment, namely computing the relative rotation between two shapes (centered at the origin and normalized by scale) by aligning spherical harmonic coefficients of their spherical function representations. Unlike most prior work, which focuses on the regime in which the inputs have approximately the same shape, we focus on the more general and challenging setting in which the shapes may differ. Central to our approach is a stability analysis of spherical harmonic coefficients, which sheds light on how to align them for robust rotation estimation. We observe that due to symmetries, certain spherical harmonic coefficients may vanish. As a result, using a robust norm for alignment that automatically discards such coefficients offers more accurate rotation estimates than the widely used L2 norm. To enable efficient continuous optimization, we show how to analytically compute the Jacobian of spherical harmonic coefficients with respect to rotations. We also introduce an efficient approach for rotation initialization that requires only a sparse set of rotation samples. Experimental results show that our approach achieves better accuracy and efficiency compared to baseline approaches.

We introduce GenAnalysis, an implicit shape generation framework that allows joint analysis of a collection of man-made shapes. GenAnalysis innovates in learning an implicit shape generator to reconstruct a continuous shape space from the input shape collection. It offers interpolations between pairs of input shapes for correspondence computation. It also allows us to understand the shape variations of each shape in the context of neighboring shapes. Such variations provide segmentation cues. A key idea of GenAnalysis is to enforce an as-affine-as-possible (AAAP) deformation regularization loss among adjacent synthetic shapes of the generator. This loss forces the generator to learn the underlying piece-wise affine part structures. We show how to extract data-driven segmentation cues by recovering piece-wise affine vector fields in the tangent space of each shape and how to use this generator to compute consistent inter-shape correspondences. These correspondences are then used to aggregate single-shape segmentation cues into consistent segmentations. Experimental results on benchmark datasets show that GenAnalysis achieves state-of-the-art results on shape segmentation and shape matching.

Vector displacement map is a common tool for artists to add parts or surface details to 3D shapes. However, creating VDM is tedious as it requires artists enormous effort to model 3D shapes, Despite the growing popularity of text/image to 3D object model, most of them has fail to generate details comparable to VDM. In this project, we present a system to automatically generate a VDM conditioned on text prompt or an image. To tackle the challenge of VDM data scarcity, we propose a multi-view normal image generation model that fineturn stable diffusion model on part-level objaverse data. We then build a reconstruction model to directly reconstruct VDM from multi-view data. Preliminary result shows that our model outperforms text/image to 3D models in detail generation.

Sketch design concepts are recurring patterns found in parametric CAD sketches.Though rarely explicitly formalized by the CAD designers, these concepts are implicitly used in design for modularity and regularity. In this paper, we propose a learning based approach that discovers the modular concepts by induction over raw sketches. We propose the dual implicit-explicit representation of concept structures that allows implicit detection and explicit generation, and the separation of structure generation and parameter instantiation for parameterized concept generation, to learn modular concepts by end-to-end training. We demonstrate the design concept learning on a large scale CAD sketch dataset and show its applications for design intent interpretation and auto-completion.

In orthodontic treatment, a full tooth model consisting of both the crown and root is indispensable in making the treatment plan. However, acquiring tooth root information to obtain the full tooth model from CBCT images is sometimes restricted due to the massive radiation of CBCT scanning. Thus, reconstructing the full tooth shape from the ready-to-use input, e.g., the partial intra-oral scan and the 2D panoramic image, is an applicable and valuable solution. In this paper, we propose a neural network, called ToothInpaintor, that takes as input a partial 3D dental model and a 2D panoramic image and reconstructs the full tooth model with high-quality root(s). Technically, we utilize the implicit rep- resentation for both the 3D and 2D inputs, and learn a latent space of the full tooth shapes. At test time, given an input, we successfully project it to the learned latent space via neural optimization to obtain the full tooth model conditioned on the input. To help find the robust projec- tion, a novel adversarial learning module is exploited in our pipeline. We extensively evaluate our method on a dataset collected from real-world clinics. The evaluation, comparison, and comprehensive ablation studies demonstrate that our approach