We introduce the simple idea of adaptive view planning to multi-view synthesis, aiming to improve both occlusion revelation and 3D consistency for single-view 3D reconstruction. Instead of producing an unordered set of views independently or simultaneously, we generate a sequence of views, leveraging temporal consistency to enhance 3D coherence. Importantly, our view sequence is not determined by a pre-determined and fixed camera setup. Instead, we compute an adaptive camera trajectory (ACT), to maximize the visibility of occluded regions of the 3D object to be reconstructed. Once the best orbit is found, we feed it to a video diffusion model to generate novel views around the orbit, which can then be passed to any multi-view 3D reconstruction model to obtain the final result. Our multiview synthesis pipeline is quite efficient since it involves no run-time training/optimization, only forward inferences by applying pre-trained models for occlusion analysis and multi-view synthesis. Our method predicts camera trajectories that reveal occlusions effectively and produce consistent novel views, significantly improving 3D reconstruction over SOTA alternatives on the unseen GSO dataset.

Nonconductive materials are extensively used in industrial applications, particularly as coatings for metal structures like oil pipelines. However, these nonmetallic coatings are prone to damage from factors, such as corrosion and scratches, leading to widespread failures. This increases the demand for nondestructive evaluation techniques capable of accurately quantifying defect parameters in such materials. Capacitive Imaging (CI) technique is an emerging electromagnetic nondestructive testing method with promising application prospects in defect evaluation in nonconducting materials. However, the CI technique is commonly used as a screening technique to detect the presence of possible defects, and its defect sizing ability, which is crucial in some engineering applications, has yet to be explored. This article proposes a high precision defect sizing method for the CI technique based on a physics informed neural network. First, the physical model of the CI technique for the detection of defects in nonconducting material is analyzed. A physical formula, which was later used as physical information, for the quantification of defect length and width was then obtained. Finite-element simulations were then conducted to visualize the sensitivity distribution of the CI sensor and analyze the characteristics of defect signals the physical information was integrated into a neural network, enabling it to quantify defect parameters from the CI detection data. Experimental results demonstrate that this method can accurately determine defect length, width, depth, and buried depth. Compared to other neural network structures and traditional algorithms, the proposed approach achieves superior precision in defect quantification.

The industrial internet of things (IIoT) necessitates robust cross-domain authentication to secure sensitive on-site equipment data. This paper presents a refined reputation-based lightweight consensus mechanism (LRBCM) tailored for IIoT's distributed network structures. Leveraging node reputation values, LRBCM streamlines ledger consensus, minimizing communication overhead and complexity. Comparative experiments show LRBCM outperforms competing mechanisms. It maintains higher throughput as the number of participating nodes increases and achieves a throughput approximately 10.78% higher than ReCon. Moreover, runtime analysis demonstrates LRBCM's scalability, surpassing ReCon by approximately 12.79% with equivalent nodes and transactions. In addition, as a combination of LRBCM, the proposed distributed lightweight authentication mechanism (ELAM) is rigorously evaluated against the security of various attacks, and its resilience is confirmed. Experiments show that ELAM has good efficiency while maintaining high security.