Xin Li

Point cloud attribute compression is challenged by fitting the attribute signals living on irregular geometric structures. Existing methods cannot achieve compact multiscale representation for high-fidelity reconstruction using the handcrafted transforms or deep learning-based techniques. In this paper, we propose a novel geometry-aware lifting-based multiscale network via spatial-channel lifting scheme for point cloud attribute compression. The proposed network cascades geometry-aware spatial lifting to reduce spatial redundancy by adaptively capturing irregular geometric structures and progressive channel lifting to progressively reduce channel-wise redundancy in multiscale representation. Furthermore, we design the split, predict, and update operations for geometry-aware spatial lifting to fully exploit the geometry information representing irregular structures. We develop geometry-aware adaptive split to equally split input points with significance scores indicating their dependencies, and propose geometry-aware cross-attention filtering for the predict and update operations for decorrelation based on geometry information. To our best knowledge, this paper achieves the first lifting-based learned transform for point cloud compression that enjoys reversibility guarantees of multiscale representation to enhance rate-distortion performance. Experimental results show that the proposed framework achieves state-of-the-art performance on extensive point cloud datasets, and outperforms latest MPEG G-PCC standard and most recent deep learning based methods.

The accurate prediction of offshore platform and ship motion is crucial for motion compensation devices and for helping the crew make informed decisions. Traditional time series and physical models are being replaced by machine learning models due to their simplicity and lower training cost. However, insufficient data has hindered model training, making evaluating and comparing different models difficult. This paper introduces a comprehensive motion dataset containing data of more than 400 pieces from tens of offshore platform tests conducted at the State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University. The dataset is divided into subsets tailored for four application scenarios, including specific types of offshore platforms, wave conditions, noise addition data, and transfer learning. A Convolutional Attention-based LSTM model that combines convolution and self-attention mechanisms is proposed to validate the dataset and improve the accuracy of motion prediction. The proposed model is compared with classical models using our introduced dataset, achieving 5–10% improvement and confirming the dataset’s high reliability and applicability, as well as the effectiveness of the Conv-Att-LSTM model. This development sets a new standard for motion prediction and furthers the application of machine learning in ocean engineering.

The emergence of polarization image sensors presents both opportunities and challenges for real-time full-polarization reconstruction in scene imaging. This paper presents an innovative three-stage interpolation method specifically tailored for monochrome polarization image demosaicking, emphasizing both precision and processing speed. The method introduces a novel linear interpolation model based on polarization channel difference priors in the initial two stages. To enhance results through bidirectional interpolation, a continuous adaptive edge detection method based on variance differences is employed for weighted averaging. In the third stage, a total intensity map, derived from the previous two stages, is integrated into a residual interpolation process, thereby further elevating estimation precision. The proposed method undergoes validation using publicly available advanced datasets, showcasing superior performance in both global parameter evaluations and local visual details when compared with existing state-of-the-art techniques.

The study on flow around a bluff body has been investigated using a variety of methods experimentally and numerically. Here we use a novel approach to obtain the flow field with excellent accuracy compared with traditional methods. Deep machine learning is combined with the issues of fluid mechanics. Flow fields around a static hydrofoil gained from computational fluid dynamics

*In general, the lifetime of the offshore oil and gas platform is about 20-30 years, and a large number of oil and gas platforms is approaching the end of their service. So, how to decommission these oil and gas platforms safely and cleanly becomes an urgent problem for the oil and gas industry and related organizations. And the costs for decommissioning equipment and services are