3D Object Classification via Spherical Projections
By Zhangjie Cao, Qi-Xing Huang, Karthik Ramani. 2017
74 citations recorded by the source corpus.
In this paper, we introduce a new method for classifying 3D objects. Our main idea is to project a 3D object onto a spherical domain centered around its barycenter and develop neural network to classify the spherical projection. We introduce two complementary projections. The first captures depth variations of a 3D object, and the second captures contour-information viewed from different angles. Spherical projections combine key advantages of two main-stream 3D classification methods: image-based and 3D-based. Specifically, spherical projections are locally planar, allowing us to use massive image datasets (e.g, ImageNet) for pre-training. Also spherical projections are similar to voxel-based methods, as they encode complete information of a 3D object in a single neural network capturing dependencies across different views. Our novel network design can fully utilize these advantages. Experimental results on ModelNet40 and ShapeNetCore show that our method is superior to prior methods.
Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Geology
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computational Mechanics
- Jiajun Wu
- Josh Tenenbaum
- Zhoutong Zhang
- Chengkai Zhang
- Xiuming Zhang
- Subhransu Maji
- Tolga Birdal
- Haowen Deng
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Ming-Hsuan Yang
- Jitendra Malik
- Dacheng Tao
- Andrea Vedaldi
- Xiaogang Wang
- Trevor Darrell
- Andrew Zisserman
In Geology
- Jiwen Lu
- Jie Zhou
- Yongming Rao
- Karthik Ramani
- Jiajun Wu
- Subhransu Maji
- Cheng Wang
- Saifullahi Aminu Bello
Citations in the corpus
- Learning to Reconstruct Shapes from Unseen Classes (2018, methodology): “, 2018], classification [Cao et al., 2017], and finding possible rotational as well as reflective symmetries [Kazhdan et al.”
- Deep Learning 3D Shapes Using Alt-az Anisotropic 2-Sphere Convolution (2018, methodology): “…methods include encoding a 3D shape as a collection of 2D renderings from multiple cameras (Qi et al. (2016); Su et al. (2015); Bai et al. (2016),) or projecting a 3D object onto geometric entities which can be flattened as 2D images (Shi et al. (2015); Cao et al. (2017); Sfikas et al. (2018).)”
- PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptors (2018, background): “Cao et al. [7] use spherical projections to aid object classification.”
- Spherical Fractal Convolutional Neural Networks for Point Cloud Recognition (2019, background): “Previous works [6, 3] discretize sphere with equiangular sampling, where the cell area varies significantly along latitude.”
- 3D2SeqViews: Aggregating Sequential Views for 3D Global Feature Learning by CNN With Hierarchical Attention Aggregation (2019, methodology): “For fair comparison, the result of VRN [31] is presented with a single CNN, where twice more views than ours are employed, the result of RotationNet [21] is presented with views taken by the default camera system orientation which keeps identical with other methods, and the result of Spherical proje”
- Flex-Convolution - Million-Scale Point-Cloud Learning Beyond Grid-Worlds (2018, unknown): “[4] propose to use a spherical projection.”
- Review: deep learning on 3D point clouds (2020, methodology): “These methods [17, 18, 19, 10, 20, 21, 22], take advantage of the already matured 2D CNNs into 3D.”
- SPLATNet: Sparse Lattice Networks for Point Cloud Processing (2018, background): “Multi-view networks pre-process shapes into a set of 2D rendered images encoding surface depth and normals under various 2D projections [39, 32, 3, 24, 9, 20].”