Slice-sampling based 3D Object Classification

Zhao Xiangwen, Yang Yi-Jun, Zeng Wei, Yang Liqun, Wang Yao
Proceedings of The 13th Asian Conference on Machine Learning, PMLR 157:689-704, 2021.

Abstract

Multiview-based 3D object detection achieved great success in the past years. However, for some complex models with complex inner structures, the performances of these methods are not satisfactory. This paper provides a method based on slide sampling for 3D object classification. First, we slice and sample the model from the different depths and directions to get the model’s features. Then, a deep neural network designed based on the attention mechanism is used to classify the input data. The experiments show that the performance of our method is competitive on ModelNet. Moreover, for some special models with simple surfaces and complex inner structures, the performance of our method is outstanding and stable.

Cite this Paper

BibTeX
@InProceedings{pmlr-v157-xiangwen21a, title = {Slice-sampling based 3D Object Classification}, author = {Xiangwen, Zhao and Yi-Jun, Yang and Wei, Zeng and Liqun, Yang and Yao, Wang}, booktitle = {Proceedings of The 13th Asian Conference on Machine Learning}, pages = {689--704}, year = {2021}, editor = {Balasubramanian, Vineeth N. and Tsang, Ivor}, volume = {157}, series = {Proceedings of Machine Learning Research}, month = {17--19 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v157/xiangwen21a/xiangwen21a.pdf}, url = {https://proceedings.mlr.press/v157/xiangwen21a.html}, abstract = {Multiview-based 3D object detection achieved great success in the past years. However, for some complex models with complex inner structures, the performances of these methods are not satisfactory. This paper provides a method based on slide sampling for 3D object classification. First, we slice and sample the model from the different depths and directions to get the model’s features. Then, a deep neural network designed based on the attention mechanism is used to classify the input data. The experiments show that the performance of our method is competitive on ModelNet. Moreover, for some special models with simple surfaces and complex inner structures, the performance of our method is outstanding and stable.} }
Endnote
%0 Conference Paper %T Slice-sampling based 3D Object Classification %A Zhao Xiangwen %A Yang Yi-Jun %A Zeng Wei %A Yang Liqun %A Wang Yao %B Proceedings of The 13th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Vineeth N. Balasubramanian %E Ivor Tsang %F pmlr-v157-xiangwen21a %I PMLR %P 689--704 %U https://proceedings.mlr.press/v157/xiangwen21a.html %V 157 %X Multiview-based 3D object detection achieved great success in the past years. However, for some complex models with complex inner structures, the performances of these methods are not satisfactory. This paper provides a method based on slide sampling for 3D object classification. First, we slice and sample the model from the different depths and directions to get the model’s features. Then, a deep neural network designed based on the attention mechanism is used to classify the input data. The experiments show that the performance of our method is competitive on ModelNet. Moreover, for some special models with simple surfaces and complex inner structures, the performance of our method is outstanding and stable.
APA
Xiangwen, Z., Yi-Jun, Y., Wei, Z., Liqun, Y. & Yao, W.. (2021). Slice-sampling based 3D Object Classification. Proceedings of The 13th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 157:689-704 Available from https://proceedings.mlr.press/v157/xiangwen21a.html.

Related Material