Unsupervised Primitive Discovery for Improved 3D Generative Modeling

Khan, Salman; Guo, Yulan; Hayat, Munawar; Barnes, Nick

doi:10.1109/cvpr.2019.00997

Cited by 35 publications

(23 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We have first evaluated our 2DSlicesNet method in classification task on the ModelNet10 and ModelNet40's test subset. We have compared our approach with recent stateof-the-art approaches, including methods that do not use machine learning; Spherical Harmonics (SPH) [23] and Light Field (LFD) [18] descriptors, and those that use machine learning, namely RadialNet [28], LonchaNet [8], Primitive-GAN [39], VSL [40], BinVoxNetPlus [41], DeepSets [42], 3D-DescriptorNet [43], the approach proposed by Soltani et al [44], the approach presented by Zanuttigh and Minto [45], ECC [46], FPNN [47], PointNet [26], PointNet [48], LightNet [49], the approach proposed by Xu and Todorovic [50], Geometry Image [51], 3D-GAN [52], Pairwise [33], MVCNN [32], GIFT [53], VoxNet [30], DeepPano [34] and 3DShapeNets [29]. Table .1 recapitulates the classification accuracy of the abovementioned approaches.…”

Section: D Object Classificationmentioning

confidence: 99%

“…ModelNet10 ModelNet40 2DSlicesNet(Ours) 94.05% 91.05% LonchaNet [8] 94.37% -RadialNet [28] 89.53% 86.42% Primitive-GAN [39] 92.20% 86.40% VSL [40] 91.00% 84.50% BinVoxNetPlus [41] 92.32% 85.47% DeepSets [42] -90.30% 3D-DescriptorNet [43] 92.40% -Soltani et al [44] -82.10% Zanuttigh and Minto [45] 91.50% 87.80% ECC [46] 90.00% 83.20% FPNN [47] -88.40% PointNet [26] -89.20% PointNet [48] 77.60% -LightNet [49] 93.94% 88.93% Xu and Todorovic [50] 88.00% 81.26% Geometry Image [51] 88.40% 83.90% 3D-GAN [52] 91.00% 83.30% Pairwise [33] 92.80% 90.70% MVCNN [32] -90.10% GIFT [53] 92.35% 83.10% VoxNet [30] 92.00% 83.00% DeepPano [34] 85.45% 77.63% 3DShapeNets [29] 83.50% 77.00% LFD (NON-ML) [18] 79.87% 75.47% SPH (NON-ML) [23] 79.79% 68.23% the misclassification originates from the similar class pairs such as dresser versus night stand and desk versus table. Fig.…”

Section: Approachesmentioning

confidence: 99%

See 1 more Smart Citation

2DSlicesNet: A 2D Slice-Based Convolutional Neural Network for 3D Object Retrieval and Classification

2021

View full text Add to dashboard Cite

3D data can be instrumental to the computer vision field as it provides insightful information about the full 3D models' geometry. Recently, with easy access to both computational power and huge 3D databases, it is feasible to apply convolutional neural networks to automatically extract the 3D models' features. This paper presents a novel approach, called 2DSlicesNet, which deals with the issue of 3D model retrieval and classification using a 2D slice-based representation with a 3D convolutional neural network.The assumption in this context is that similar 3D models will be composed of almost identical 2D slices. Therefore, we first transform each normalized 3D model into a set of 2D slices corresponding to its first main axis, and then use them as input data to our 3D convolutional neural network. Experimental results and comparison with state-of-the-art approaches, using ModelNet10 and ModelNet40 datasets, prove that our proposed 2DSlicesNet approach can reach notable rates of accuracy in classification and retrieval.INDEX TERMS Deep learning, 2D slices, 3D Convolutional neural network, 3D object classification, 3D object retrieval.

show abstract

Section: D Object Classificationmentioning

confidence: 99%

Section: Approachesmentioning

confidence: 99%

2DSlicesNet: A 2D Slice-Based Convolutional Neural Network for 3D Object Retrieval and Classification

2021

View full text Add to dashboard Cite

show abstract

“…The best average class accuracy of 91.4% was achieved using multiorientation pooling strategy To reduce the computational cost of voxel representation, NormalNet [10] and LP-3DCNN [33] reduced the number of parameters by introducing RCC and ReLPV modules respectively with better classification accuracy than state-of-art-methods. Recently, Khan et al proposed an unsupervised primitive GAN [34] model, where generated volumetric features were used to train the network on ModelNet10 dataset for the classification task but tested on ModelNet10 and ModelNet40 dataset. The accuracy of this unsupervised method is inferior to supervised methods.…”

Section: A Deep Neural Network For 3d Object Classificationmentioning

confidence: 99%

3D Object Classification Using a Volumetric Deep Neural Network: An Efficient Octree Guided Auxiliary Learning Approach

et al. 2020

View full text Add to dashboard Cite

We consider the recent challenges of 3D shape analysis based on a volumetric CNN that requires a huge computational power. This high-cost approach forces to reduce the volume resolutions when applying 3D CNN on volumetric data. In this context, we propose a multiorientation volumetric deep neural network (MV-DNN) for 3D object classification with octree generating low-cost volumetric features. In comparison to conventional octree representations, we propose to limit the octree partition to a certain depth to reserve all leaf octants with sparsity features. This allows for improved learning of complex 3D features and increased prediction of object labels at both low and high resolutions. Our auxiliary learning approach predicts object classes based on the subvolume parts of a 3D object that improve the classification accuracy compared to other existing 3D volumetric CNN methods. In addition, the influence of views and depths of the 3D model on the classification performance is investigated through extensive experiments applied to the ModelNet40 database. Our deep learning framework runs significantly faster and consumes less memory than full voxel representations and demonstrate the effectiveness of our octree-based auxiliary learning approach for exploring high resolution 3D models. Experimental results reveal the superiority of our MV-DNN that achieves better classification accuracy compared to state-of-art methods on two public databases. INDEX TERMS 3D shape analysis, object classification, convolutional neural network, DNNs, volumetric CNN. A. A. M. MUZAHID received the M.E. degree in communication and information engineering from the Chongqing University of Posts and Telecommunications, Chongqing, China, in 2016, and the B.Sc. degree in electronics and telecommunications engineering from Daffodil International University, Dhaka, Bangladesh, in 2011. He is currently pursuing the Ph.D. degree in communication and information systems with Shanghai University, Shanghai, China. He is a Research Member of the Institute of Smart City, Shanghai University, and leading the ''Computer Vision and 3D Virtual Reality'' research team. His current research interests include 3D Shape analysis, computer vision, and 3D virtual reality.

show abstract

“…The standard deconvolutional layer is not the only building block which the generator can be constructed with. A novel factorized generative model in [17] generates the 3D models by first building it with generic primitives and then refining the resulting shape. The resulting representation was evaluated using a classifier which achieved state-of-the-art results on ModelNet10 among the unsupervised models.…”

Section: Related Workmentioning

confidence: 99%

Generative Enhancement of 3D Image Classifiers

Varga

Jadlovsky

Jadlovska

2020

Preprint

View full text Add to dashboard Cite

In this paper, we propose a methodology for generative enhancement of existing 3D image classifiers. This methodology is based on combining the strengths of both non-generative classifiers and generative modeling. Its purpose is to streamline the creation of new classifiers by embedding existing compatible classifiers in a generative network architecture. The demonstration of this process and evaluation of its effects is performed using a 3D convolutional classifier and its generative equivalent - a conditional generative adversarial network classifier. The results show that the generative model achieves greater classification performance, gaining a relative classification accuracy improvement of 7.43%. Improvement of accuracy is also present when compared to a plain convolutional classifier trained on a dataset augmented with examples produced by a trained generator. This suggests there is a desirable knowledge sharing within the hybrid discriminator-classifier network.

show abstract

Unsupervised Primitive Discovery for Improved 3D Generative Modeling

Cited by 35 publications

References 39 publications

2DSlicesNet: A 2D Slice-Based Convolutional Neural Network for 3D Object Retrieval and Classification

2DSlicesNet: A 2D Slice-Based Convolutional Neural Network for 3D Object Retrieval and Classification

3D Object Classification Using a Volumetric Deep Neural Network: An Efficient Octree Guided Auxiliary Learning Approach

Generative Enhancement of 3D Image Classifiers

Contact Info

Product

Resources

About