More Like this

1. SurfGen: Adversarial 3D Shape Synthesis with Explicit Surface Discriminators

   Luo, Andrew; Li, Tianqin; Zhang, Wen-Hao; Lee, Tai-Sing (October 2021, Proceedings)

   null (Ed.)

   Recent advances in deep generative models have led to immense progress in 3D shape synthesis. While existing models are able to synthesize shapes represented as voxels, point-clouds, or implicit functions, these methods only indirectly enforce the plausibility of the final 3D shape surface. Here we present a 3D shape synthesis framework (SurfGen) that directly applies adversarial training to the object surface. Our approach uses a differentiable spherical projection layer to capture and represent the explicit zero isosurface of an implicit 3D generator as functions defined on the unit sphere. By processing the spherical representation of 3D object surfaces with a spherical CNN in an adversarial setting, our generator can better learn the statistics of natural shape surfaces. We evaluate our model on large-scale shape datasets, and demonstrate that the end-to-end trained model is capable of generating high fidelity 3D shapes with diverse topology. 

   more » « less
2. Multi-level 3D CNN for Learning Multi-scale Spatial Features

   Ghadai, Sambit; Yeow Lee, Xian; Balu, Aditya; Sarkar, Soumik; Krishnamurthy, Adarsh (January 2019, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops)

   3D object recognition accuracy can be improved by learning the multi-scale spatial features from 3D spatial geometric representations of objects such as point clouds, 3D models, surfaces, and RGB-D data. Current deep learning approaches learn such features either using structured data representations (voxel grids and octrees) or from unstructured representations (graphs and point clouds). Learning features from such structured representations is limited by the restriction on resolution and tree depth while unstructured representations creates a challenge due to non-uniformity among data samples. In this paper, we propose an end-to-end multi-level learning approach on a multi-level voxel grid to overcome these drawbacks. To demonstrate the utility of the proposed multi-level learning, we use a multi-level voxel representation of 3D objects to perform object recognition. The multi-level voxel representation consists of a coarse voxel grid that contains volumetric information of the 3D object. In addition, each voxel in the coarse grid that contains a portion of the object boundary is subdivided into multiple fine-level voxel grids. The performance of our multi-level learning algorithm for object recognition is comparable to dense voxel representations while using significantly lower memory. 

   more » « less
3. Fast Non-Convex Hull Computation

   https://doi.org/10.1109/3DV.2019.00087  

   Bayardo Spadafora, Julian; Gomez-Fernandez, Francisco; Taubin, Gabriel (September 2019, 2019 International Conference on 3D Vision (3DV))

   3D surface reconstruction usually begins with a point cloud and aims to build a representation of the object producing that point cloud. There are several algorithms to solve this problem, each with different priors over the point cloud, such as the type of object represented, or the method by which it was obtained. In this work, we focus on an algorithm called Non-Convex Hull (NCH), which reconstructs surfaces through a concept similar to the Medial Axis Transform. A new algorithm called Shrinking Planes is proposed to compute the NCH, based on the Shrinking Ball method with a few improvements. We prove that the new method can approximate surfaces to arbitrarily small error, and evaluate its performance on the surface reconstruction task. The new method maintains the same reconstruction quality as the Naïve Non-Convex Hull method, while achieving a large performance improvement. 

   more » « less
4. Synthesizing 3D Shapes From Silhouette Image Collections Using Multi-Projection Generative Adversarial Networks

   https://doi.org/10.1109/CVPR.2019.00568  

   Li, Xiao; Dong, Yue; Peers, Pieter; Tong, Xin (June 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR))

   We present a new weakly supervised learning-based method for generating novel category-specific 3D shapes from unoccluded image collections. Our method is weakly supervised and only requires silhouette annotations from unoccluded, category-specific objects. Our method does not require access to the object's 3D shape, multiple observations per object from different views, intra-image pixel correspondences, or any view annotations. Key to our method is a novel multi-projection generative adversarial network (MP-GAN) that trains a 3D shape generator to be consistent with multiple 2D projections of the 3D shapes, and without direct access to these 3D shapes. This is achieved through multiple discriminators that encode the distribution of 2D projections of the 3D shapes seen from a different views. Additionally, to determine the view information for each silhouette image, we also train a view prediction network on visualizations of 3D shapes synthesized by the generator. We iteratively alternate between training the generator and training the view prediction network. We validate our multi-projection GAN on both synthetic and real image datasets. Furthermore, we also show that multi-projection GANs can aid in learning other high-dimensional distributions from lower dimensional training datasets, such as material-class specific spatially varying reflectance properties from images. 

   more » « less
5. CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations

   Rempe, Davis; Birdal, Tolga; Zhao, Yongheng; Gojcic, Zan; Sridhar, Srinath; Guibas, Leonidas (January 2020, Advances in neural information processing systems)

   null (Ed.)

   We propose CaSPR, a method to learn object-centric Canonical Spatiotemporal Point Cloud Representations of dynamically moving or evolving objects. Our goal is to enable information aggregation over time and the interrogation of object state at any spatiotemporal neighborhood in the past, observed or not. Different from previous work, CaSPR learns representations that support spacetime continuity, are robust to variable and irregularly spacetime-sampled point clouds, and generalize to unseen object instances. Our approach divides the problem into two subtasks. First, we explicitly encode time by mapping an input point cloud sequence to a spatiotemporally-canonicalized object space. We then leverage this canonicalization to learn a spatiotemporal latent representation using neural ordinary differential equations and a generative model of dynamically evolving shapes using continuous normalizing flows. We demonstrate the effectiveness of our method on several applications including shape reconstruction, camera pose estimation, continuous spatiotemporal sequence reconstruction, and correspondence estimation from irregularly or intermittently sampled observations. 

   more » « less