GRNet: Gridding Residual Network for Dense Point Cloud Completion

Xie, Haozhe; Yao, Hongxun; Zhou, Shangchen; Mao, Jiageng; Zhang, Shengping; Sun, Wenxiu

doi:10.1007/978-3-030-58545-7_21

Cited by 254 publications

(262 citation statements)

References 40 publications

Supporting

Mentioning

255

Contrasting

Unclassified

Order By: Relevance

“…to break the uniform plane topology and make the patch points closer to the ground truth. Unlike GRNet [18] transforms point clouds to 3D grids as intermediate representations so that they can apply 3D CNN to generate coarse points, our work introduces 3D grid points as extra auxiliary points to restrain the generation of fine points. We still treat the 3D grid points as point clouds.…”

Section: D Grid Transformation Networkmentioning

confidence: 99%

3D Grid Transformation Network For Point Cloud Completion

Deng

Buris

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Point cloud completion is of significant importance in 3D scene modeling because point clouds collected from sensors tend to be sparse and incomplete. Existing methods prefer to transform 2D grid points to the underlying surface of output point clouds. However, the point clouds they complete commonly retain the original plane structure and fail to recover details. To solve the problem, we propose 3D Grid Transformation Network. Unlike transforming 3D grid points to point clouds directly, we calculate their weights for the reconstructed point clouds. Our method can break the topology of the 3D grid structure and make the underlying shape of output point clouds closer to the ground truth. We validate our method on point cloud completion and experimental results show that our method can achieve better quantitative results and obtain fine-grained details.

show abstract

Section: D Grid Transformation Networkmentioning

confidence: 99%

3D Grid Transformation Network For Point Cloud Completion

Deng

Buris

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…Görüntü elde etme teknolojilerindeki yeni gelişmeler ve 3B sensörlerin yaygın kullanımıyla, 3B nesne işleme algoritmalarına olan talep de artmıştır. Bu amaçla son zamanlarda çesitli algoritmalar geliştirilmiştir [1]. Bu alanlardan birisi de 3B tabanlı nokta bulutlarının işlenmesidir.…”

Section: G İr İ şunclassified

GRJointNET: Synergistic Completion and Part Segmentation on 3D Incomplete Point Clouds

Yigit

Taspinar

Yurt

et al. 2021

2021 29th Signal Processing and Communications Applications Conference (SIU)

View full text Add to dashboard Cite

Özetçe -Üç boyutlu (3B) nokta bulutları üzerinde bölütleme yapmak, otonom sistemler için önemli ve gerekli bir işlemdir. Bölütleme algoritmalarının başarısı, üzerinde işlem yapılan nokta bulutlarının niteligine (çözünürlük, tamlık vb.) baglıdır. Dolayısıyla, nokta bulutundaki mevcut eksiklikler, nokta bulutu tabanlı uygulamaların başarısını düşürmektedir. Bu konuda, güncel bir çalısma olan GRNet, eksik nokta bulutlarını tamamlamaya odaklanan başarılı bir algoritmadır, ancak bölütleme yetenegi yoktur. Biz bu çalışmada, GRNet üzerine geliştirdigimiz derin ögrenme tabanlı GRJointNet algoritmasını sunmaktayız. GRJointNet hem bir nokta bulutundaki eksik noktaları tamamlamakta, hem de onun yapamadıgı parça bölütlemesini de yapmaktadır. Bu işlemler elde ettikleri verileri birbirlerini desteklemek için kullanmaktadır. ShapeNet-Part veri kümesinde yapılmış deneylerimiz, GR-JointNet algoritmasının nokta bulutu tamamlamada GRNet'den daha başarılı oldugunu göstermektedir. Aynı zamanda, GRNet bölütleme yapamazken, GRJointNet bu özelligi de kazanmıştır. Dolayısıyla bu çalışma nokta bulutlarının 3B bilgisayarlı görüde kullanışlılıgını arttırmak adına umut vadetmektedir.

show abstract

“…Furthermore, to revive the structure information loss in the DNN, some of the methods reconstruct the filled point cloud of an object by computing the collection of parametric surface elements [31]. GRNet [1] addressed this issue by using the 3D convolution neural network with gridding loss function to get the coarse to fine-scale point clouds. Instead of using the 3D convolutional layer to reduce the high cost and memory requirement, we use the MLP based framework along with an auto-encoder, and tree-based decoder [11].…”

Section: Related Workmentioning

confidence: 99%

“…To restore the complete point cloud from its partial version is called the point cloud or shape completion task. This finds its application in tasks such as semantic segmentation [1], shape classification [2], point cloud registration [3] and SLAM [4].…”

Section: Introductionmentioning

confidence: 99%

3d Point Cloud Completion Using Stacked Auto-Encoder For Structure Preservation

Kumari

Raman

2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

3D point cloud completion problem deals with completing the shape from partial points. The problem finds its application in many visionrelated applications. Here, structure plays an important role. Most of the existing approaches either do not consider structural information or consider structure at the decoder only. For maintaining the structure, it is also necessary to maintain the position of the available 3D points. However, most of the approaches lack the aspect of maintaining the available structural position. In this paper, we propose to employ stacked auto-encoder in conjunction a with shared Multi-Layer Perceptron (MLP). MLP converts each 3D point into a feature vector and the stacked auto-encoder helps in maintaining the available structural position of the input points. Further, it explores the redundancy present in the feature vector. It aids to incorporate coarse to fine scale information that further helps in better shape representation. The embedded feature is finally decoded by a structural preserving decoder. Both the encoding and the decoding operations of our method take care of preserving the structure of the available shape information. The experimental results demonstrate the structure preserving capability of our network as compared to the stateof-the-art methods.

show abstract

GRNet: Gridding Residual Network for Dense Point Cloud Completion

Cited by 254 publications

References 40 publications

3D Grid Transformation Network For Point Cloud Completion

3D Grid Transformation Network For Point Cloud Completion

GRJointNET: Synergistic Completion and Part Segmentation on 3D Incomplete Point Clouds

3d Point Cloud Completion Using Stacked Auto-Encoder For Structure Preservation

Contact Info

Product

Resources

About