相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Computer Science, Software Engineering
Chun-Yu Sun et al.
Summary: The lack of fine-grained 3D shape segmentation data is a major challenge for developing learning-based 3D segmentation techniques. In this study, we propose an effective semi-supervised method that learns 3D segmentations using a combination of labeled and unlabeled data. Our approach incorporates a novel multilevel consistency loss to ensure consistent network predictions across different levels of perturbed 3D shapes, and a part substitution scheme to augment labeled data for improved training. Extensive validation on different tasks demonstrates the superior performance of our method compared to existing semi-supervised and unsupervised pre-training approaches.
COMPUTATIONAL VISUAL MEDIA
(2023)
Article
Computer Science, Artificial Intelligence
Haonan Yang et al.
Summary: In this paper, a pyramid structure network combining a convolutional neural network and Swin Transformer is proposed for automatic segmentation of breast tumors. The network improves the performance of breast lesion segmentation by using interactive channel attention and supplementary feature fusion modules, and introduces a boundary detection module to enhance the boundary quality of segmentation results. Experimental results demonstrate the superiority of the network in breast ultrasound lesion segmentation.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Review
Computer Science, Artificial Intelligence
Jun Li et al.
Summary: This paper reviews the research progress and applications of Transformers in medical imaging. After comparing Transformers with convolutional neural networks, the key defining properties of Transformers and the combination of Transformers and CNNs are highlighted. The latest research progress in medical image segmentation, recognition, detection, registration, reconstruction, enhancement, etc. are comprehensively reviewed, and the organization of the review based on key defining properties and architectural type provides the readers with the best understanding of the rationale behind the reviewed approaches.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Environmental Sciences
Yan Zhang et al.
Summary: We propose a new multi-sensor fusion detection framework called PMPF for 3D object detection in autonomous driving, which addresses the issues of high latency and coupling in sequence-based data-level fusion frameworks and performs better than LiDAR-only detectors. PMPF projects point cloud data onto the image plane, processes region pixels to correspond with the points, and decorates them to the point cloud data, enabling the fused point cloud data to be used in LiDAR-only detectors with autoencoders. PMPF is a plug-and-play, decoupled multi-sensor fusion detection framework with low latency.
Article
Computer Science, Artificial Intelligence
Fahad Shamshad et al.
Summary: This survey reviews the applications of Transformers in medical imaging, covering tasks such as medical image segmentation, detection, classification, restoration, synthesis, registration, and clinical report generation. The challenges and solutions for each application are discussed, and future research directions are highlighted. The survey aims to spark interest in the academic community and provide researchers with an up-to-date reference regarding the applications of Transformer models in medical imaging.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Computer Science, Artificial Intelligence
Javaria Amin et al.
Summary: This research proposes a novel secure framework for brain tumor diagnosis, which involves encrypting and decrypting MRI data to protect patient privacy, as well as using a quantum model and a semantic segmentation model for tumor classification and segmentation. The proposed model shows better performance compared to the latest research works.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Leyuan Liu et al.
Summary: This work proposes a deep neural network called FuDNN for LiDAR-camera fusion 3D object detection in autonomous driving. The network includes a 2D backbone and an attention-based fusion sub-network, as well as the use of RPN and PointRCNN for 3D box predictions. Experiments show that this approach achieves high performance on the KITTI dataset.
Article
Robotics
Song Wang et al.
Summary: LiDAR sensor is crucial for the perception system in autonomous vehicles and intelligent robots. This study proposes a novel approach Meta-RangeSeg for semantic segmentation in LiDAR sequences, by introducing a new range residual image representation to capture spatial-temporal information and employing Meta-Kernel and U-Net backbone for feature extraction.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Haoxi Ran et al.
Summary: In this paper, a novel representation method called RepSurf is proposed to explicitly depict the very local structure of point clouds. It achieves better performance and efficiency than previous state-of-the-art methods in classification, segmentation, and detection tasks.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Corentin Sautier et al.
Summary: This paper presents a self-supervised pretraining method tailored for autonomous driving data, utilizing synchronized image and Lidar sensors for model training. By matching features of visually similar regions using superpixels, the method improves the performance of 3D semantic segmentation and object detection tasks.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Mohamed Afham et al.
Summary: This paper presents a simple cross-modal contrastive learning approach called CrossPoint for learning transferable 3D point cloud representations. Experimental results show that this method outperforms previous unsupervised learning methods in tasks such as 3D object classification and segmentation.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2022)
Review
Computer Science, Artificial Intelligence
Guoguang Du et al.
Summary: This paper provides a comprehensive overview of vision-based robotic grasping, focusing on key tasks such as object localization, object pose estimation, and grasp estimation. Various methods combining these tasks, including traditional and deep learning-based approaches, are reviewed in detail. Challenges and future directions in the field are also highlighted.
ARTIFICIAL INTELLIGENCE REVIEW
(2021)
Article
Computer Science, Artificial Intelligence
Yulan Guo et al.
Summary: Recently, deep learning on point clouds has been gaining more attention, but it is still in its early stages due to the unique challenges faced in processing point clouds with deep neural networks.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Duarte Fernandes et al.
Summary: Autonomous vehicles are relying on deep learning and 3D scanners for better perception and object detection; since 2010, there has been a significant increase in the number and innovation of methods for self-driving systems; there is notable development in newer techniques adapted to LiDAR data.
INFORMATION FUSION
(2021)
Review
Computer Science, Hardware & Architecture
Shaohua Qi et al.
Summary: This paper comprehensively reviews and classifies the latest developments in deep learning methods for multi-view 3D object recognition, summarizes results on mainstream datasets, provides insightful conclusions, and proposes enlightening future research directions.
Article
Environmental Sciences
Jie Wan et al.
Summary: The DGANet is a Dilated Graph Attention-based Network designed to extract local geometric features by establishing local dilated graph-like regions and integrating dilated graph attention modules.
Proceedings Paper
Computer Science, Artificial Intelligence
Ishan Misra et al.
Summary: 3DETR is an end-to-end Transformer based object detection model for 3D point clouds, requiring minimal modifications to the vanilla Transformer block and outperforming specialized architectures on the ScanNetV2 dataset. It is applicable to tasks beyond detection and can serve as a building block for future research.
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Haoxi Ran et al.
Summary: This paper explores the possibilities of local relation operators and proposes a scalable and efficient module called group relation aggregator. The study shows that wider RPNet is suitable for classification, while deeper RPNet is better for segmentation. RPNet achieves state-of-the-art results for both classification and segmentation tasks.
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ye Chen et al.
Summary: The Shape Self-Correction method uses self-supervised learning to improve shape representation in point cloud analysis by distorting and correcting shapes. Experimental results show that it outperforms other unsupervised methods on various datasets and can boost the performance of supervised models in point cloud analysis.
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021)
(2021)
Proceedings Paper
Automation & Control Systems
Xuemeng Yang et al.
Summary: An end-to-end semantic segmentation-assisted scene completion network is proposed for outdoor scene completion, providing competitive performance and low latency in 3D scene understanding.
2021 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2021)
Article
Computer Science, Information Systems
Nico Engel et al.
Summary: Point Transformer is a deep neural network that operates directly on unordered and unstructured point sets, extracting local and global features and relating them through a local-global attention mechanism. SortNet induces input permutation invariance by selecting points based on a learned score. The output is a sorted and permutation invariant feature list that can be directly incorporated into common computer vision applications, showing competitive results compared to prior work through evaluation on standard benchmarks.
Proceedings Paper
Computer Science, Artificial Intelligence
Hui Cao et al.
MULTIMEDIA MODELING (MMM 2020), PT II
(2020)
Article
Chemistry, Analytical
Yan Yan et al.
Proceedings Paper
Computer Science, Artificial Intelligence
Angela Dai et al.
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
(2017)
