Results 31 to 40 of about 6,559,013 (333)
Rotation-Invariant Transformer for Point Cloud Matching [PDF]
Computer Vision and Pattern Recognition, 2023 The intrinsic rotation invariance lies at the core of matching point clouds with handcrafted descriptors. However, it is widely despised by recent deep matchers that obtain the rotation invariance extrinsically via data augmentation. As the finite number
Hao Yu +6 more
semanticscholar +1 more source
AdaPoinTr: Diverse Point Cloud Completion With Adaptive Geometry-Aware Transformers [PDF]
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023 In this paper, we propose a Transformer encoder-decoder architecture, called PoinTr, which reformulates point cloud completion as a set-to-set translation problem and employs a geometry-aware block to model local geometric relationships explicitly.
Xumin Yu +4 more
semanticscholar +1 more source
FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer [PDF]
Computer Vision and Pattern Recognition, 2023 Transformer, as an alternative to CNN, has been proven effective in many modalities (e.g., texts and images). For 3D point cloud transformers, existing efforts focus primarily on pushing their accuracy to the state-of-the-art level.
Zhijian Liu +4 more
semanticscholar +1 more source
CLIP2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Data [PDF]
Computer Vision and Pattern Recognition, 2023 Contrastive Language-Image Pre-training, benefiting from large-scale unlabeled text-image pairs, has demonstrated great performance in open-world vision understanding tasks.
Yi Zeng +9 more
semanticscholar +1 more source
Attention-Based Point Cloud Edge Sampling [PDF]
Computer Vision and Pattern Recognition, 2023 Point cloud sampling is a less explored research topic for this data representation. The most commonly used sampling methods are still classical random sampling and farthest point sampling.
Chengzhi Wu +3 more
semanticscholar +1 more source
End-to-End Point Cloud Completion Network with Attention Mechanism
Sensors, 2022 We propose a conceptually simple, general framework and end-to-end approach to point cloud completion, entitled PCA-Net. This approach differs from the existing methods in that it does not require a “simple” network, such as multilayer perceptrons (MLPs),
Yaqin Li +4 more
doaj +1 more source
Open-Vocabulary Point-Cloud Object Detection without 3D Annotation [PDF]
Computer Vision and Pattern Recognition, 2023 The goal of open-vocabulary detection is to identify novel objects based on arbitrary textual descriptions. In this paper, we address open-vocabulary 3D point-cloud detection by a dividing-and-conquering strategy, which involves: 1) developing a point ...
Yuheng Lu +6 more
semanticscholar +1 more source
A partial overlapping point cloud registration method based on dynamic feature matching
Dianxin kexue, 2021 The point cloud registration method can effectively complete the registration of point clouds with different overlap rates and various sizes, and ensure the accuracy of the 3D reconstruction model.To address the above issues, a partial overlapping point ...
Hui DU +3 more
doaj +2 more sources
Real3D-AD: A Dataset of Point Cloud Anomaly Detection [PDF]
Neural Information Processing Systems, 2023 High-precision point cloud anomaly detection is the gold standard for identifying the defects of advancing machining and precision manufacturing. Despite some methodological advances in this area, the scarcity of datasets and the lack of a systematic ...
Jiaqi Liu +7 more
semanticscholar +1 more source
Grad-PU: Arbitrary-Scale Point Cloud Upsampling via Gradient Descent with Learned Distance Functions [PDF]
Computer Vision and Pattern Recognition, 2023 Most existing point cloud upsampling methods have roughly three steps: feature extraction, feature expansion and 3D coordinate prediction. However, they usually suffer from two critical issues: (1) fixed upsampling rate after one-time training, since the
Yun He +4 more
semanticscholar +1 more source