Multi-view Harmonized Bilinear Network for 3D Object Recognition

By Tan Yu, Jingjing Meng, Junsong Yuan. 2018

299 citations recorded by the source corpus.

View-based methods have achieved considerable success in 3D object recognition tasks. Different from existing view-based methods pooling the view-wise features, we tackle this problem from the perspective of patches-to-patches similarity measurement. By exploiting the relationship between polynomial kernel and bilinear pooling, we obtain an effective 3D object representation by aggregating local convolutional features through bilinear pooling. Meanwhile, we harmonize different components inherited in the bilinear feature to obtain a more discriminative representation. To achieve an end-to-end trainable framework, we incorporate the harmonized bilinear pooling as a layer of a network, constituting the proposed Multi-view Harmonized Bilinear Network (MHBN). Systematic experiments conducted on two public benchmark datasets demonstrate the efficacy of the proposed methods in 3D object recognition.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data (2019, methodology): “, using multiple view images [39, 48, 46, 22], or applying convolutions on 3D voxel grids [27, 43].”
PF-Net: Point Fractal Network for 3D Point Cloud Completion (2020, background): “However, multi-view and voxel-based representation[18, 7, 34, 11, 13, 37] leads to unnecessarily voluminous and limits the output resolution of voxels.”
MVTN: Multi-View Transformation Network for 3D Shape Recognition (2020, background): “Several follow-up works propose different strategies to assign weights to views to perform weighted average pooling of view-specific features [74, 72, 18, 9].”
Semantically Tied Paired Cycle Consistency for Zero-Shot Sketch-Based Image Retrieval (2019, background): “in terms of memory [62], and extracting useful information from this high dimensional space could result in information loss [61].”
Deep multi-view learning methods: A review (2021, methodology): “For the deep multi-view object recognition methods, we compare the classification accuracy (Accuracy) [49] of the following 660 approaches: multi-view CNN (MVCNN) [49], Kd-tree based network (KD-Net) [51], multi-view harmonized bilinear network (MHBN) [52], group-view CNN (GVCNN) [50] and dynamic ro”
View-GCN: View-Based Graph Convolutional Network for 3D Shape Analysis (2020, unknown): “MHBN [55] and RN [53] aggregate features of multi-view patches by harmonized bilinear pooling and attention.”
Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline (2021, methodology): “3D shape Analysis using Rendered Images and Voxels: Many works use images rendered from object meshes for 3D shape analysis (Maturana & Scherer, 2015; Wu et al., 2015b; Yu et al., 2018; Guo et al., 2016; Shi et al., 2015; Hackel et al., 2017; Song & Xiao, 2016; 2014; Huang & You, 2016; Tchapmi et al”