A GRAPH-CNN FOR 3D POINT CLOUD CLASSIFICATION
提出了一種基於圖卷積神經網絡提取點特徵的點雲分類網絡。 輸入N*3的點雲，通過切比雪夫多項式的圖卷積，先升成N*1000，也就是把每個節點的信息增加到1000維，然後針對global-pooling，再做一次卷積，把兩次卷積得到的2個N*1000維的張量，按列取最大和按列算方差（對稱函數），則消除了點序的影響，然後兩個2*1000的最大向量concat，兩個2*1000的方差向量concat，得到
相關文章
- 1. 3D Shape Classification——Point-based Graph based
- 2. 讀論文 3D Point Cloud Classification using 3D Modified Fisher Vector Representation for CNN
- 3. [Point Cloud] PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation 論文解讀
- 4. 3D 點雲識別: Geometric Feedback Network for Point Cloud Classification
- 5. Point Encoder GAN: A deep learning model for 3D point cloud inpainting
- 6. 3D Shape Classification——Point-based Continuous Convolution Networks
- 7. 3D點雲網絡：PointNet:Deep Learning on Point Sets for 3D Classification and Segmentation
- 8. 04 PointAugment: an Auto-Augmentation Framework for Point Cloud Classification
- 9. Point Cloud 2019
- 10. 3D Shape Classification——Point-based Discrete Convolution Network
- 更多相關文章...
- • RSS
元素 - RSS 教程 - • PHP vprintf() 函數 - PHP參考手冊
- • Spring Cloud 微服務實戰(三) - 服務註冊與發現
- • Docker容器實戰(一) - 封神Server端技術
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. 3D Shape Classification——Point-based Graph based
- 2. 讀論文 3D Point Cloud Classification using 3D Modified Fisher Vector Representation for CNN
- 3. [Point Cloud] PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation 論文解讀
- 4. 3D 點雲識別: Geometric Feedback Network for Point Cloud Classification
- 5. Point Encoder GAN: A deep learning model for 3D point cloud inpainting
- 6. 3D Shape Classification——Point-based Continuous Convolution Networks
- 7. 3D點雲網絡：PointNet:Deep Learning on Point Sets for 3D Classification and Segmentation
- 8. 04 PointAugment: an Auto-Augmentation Framework for Point Cloud Classification
- 9. Point Cloud 2019
- 10. 3D Shape Classification——Point-based Discrete Convolution Network