Estimating the equilibrium parameters of environmental systems is a fundamental task, yet it is often hampered by sparse and noisy sensor data. While the standard Maximum Likelihood Estimator (MLE) is intuitive, Stein's paradox famously shows that it is statistically inefficient in high dimensions. To address this, we introduce STEADY, an estimator that generalises Stein's paradox by integrating physical knowledge. We derive our estimator from a principled empirical Bayes model where the prior distribution over the equilibria is a direct consequence of the stationary properties of the system's governing differential equations. This leads to a novel adaptive shrinkage mechanism, where the amount of shrinkage applied to each observation is naturally modulated by the physical stability of the measured system. We provide a rigorous frequentist analysis of our estimator, proving that STEADY not only dominates the MLE but is also minimax under certain conditions, offering the strongest possible guarantee of robustness. We validate our claims on synthetic data and demonstrate STEADY's utility on the global Argo ocean float dataset, showing that it effectively filters noise to reveal the ''North Atlantic Warming Hole''. The source code is available at https://github.com/shanfenghu/steady

Abstract During steel production, ensuring the integrity of the product’s surface is crucial for maintaining a competitive edge. Detecting surface flaws is a key component in preserving high standards of production quality, affecting both the end product’s reliability and the efficiency of the manufacturing process. Traditional methods for identifying these defects have relied heavily on manual inspection or basic computer vision techniques, both of which present a significant challenge in terms of accuracy and the health and safety of the inspectors. This paper explores an innovative approach by integrating a sequence generation model incorporating the transformer architecture, enhancing the detection of defects in the production of quality hot-rolled steel sheets. This method, which solves object detection through the lens of sequence generation, offers a more intricate analysis of images, enabling a comprehensive examination of surface anomalies along with detailed annotations regarding their nature and precise location. The approach not only proposes to increase the accuracy of defect detection but also highlights its adaptability for broader industrial uses.

The objective of human pose estimation is to accurately identify the positions of individual body joints for each person seen in an image for human kinematics modelling and analysis. A majority of current deep-learning approaches focus on utilizing feature learning to regress the coordinates of each individual joint. This can be thought of as a constrained point distribution optimization problem in the image plane. However, in the high-dimensional feature space, the feature distribution of joints as a significant constraint and discriminative condition, has not received enough attention yet. Here we propose a novel approach (MIMIC-Pose) that applies an implicit pair-wise keypoint membership constraint using those features to individual joint regression to ensure that the joints of the same person have higher mutual feature similarities compared with those of different people or the image backgrounds. We propose a novel link prediction integrated learning architecture with high-throughput tokens to learn such similarity feature embeddings to improve the prediction accuracy of individual joints in a single forward inference. Experimental results show that our approach can achieve competitive performance with much fewer model parameters and lower computational cost compared to state-of-the-art methods.

In this paper, we propose a deep metric for unifying the representation of mesh saliency detection and non-rigid shape matching. While saliency detection and shape matching are two closely related and fundamental tasks in shape analysis, previous methods approach them separately and independently, failing to exploit their mutually beneficial underlying relationship. In view of the existing gap between saliency and matching, we propose to solve them together using a unified metric representation of surface meshes. We show that saliency and matching can be rigorously derived from our representation as the principal eigenvector and the smoothed Laplacian eigenvectors respectively. Learning the representation jointly allows matching to improve the deformation-invariance of saliency while allowing saliency to improve the feature localization of matching. To parameterize the representation from a mesh, we also propose a deep recurrent neural network (RNN) for effectively integrating multi-scale shape features and a soft-thresholding operator for adaptively enhancing the sparsity of saliency. Results show that by jointly learning from a pair of saliency and matching datasets, matching improves the accuracy of detected salient regions on meshes, which is especially obvious for small-scale saliency datasets, such as those having one to two meshes. At the same time, saliency improves the accuracy of shape matchings among meshes with reduced matching errors on surfaces.

The prior knowledge of real human body shapes and poses is fundamental in computer games and animation (e.g. performance capture). Linear subspaces such as the popular SMPL model have a limited capacity to represent the large geometric variations of human shapes and poses. What is worse is that random sampling from them often produces non-realistic humans because the distribution of real humans is more likely to concentrate on a non-linear manifold instead of the full subspace. Towards this problem, we propose to learn human shape and pose manifolds using a more powerful deep generator network, which is trained to produce samples that cannot be distinguished from real humans by a deep discriminator network. In contrast to previous work that learn both the generator and discriminator in the original geometry spaces, we learn them in the more representative latent spaces discovered by a shape and a pose auto-encoder network respectively. Random sampling from our priors produces higher-quality human shapes and poses. The capacity of our priors is best applied to applications such as virtual human synthesis in games.

To address security problem of the Ad-hoc network caused by vulnerable username and password, we present a trustworthiness evaluation model based on Bayesian network. The node's trustworthiness is computed in the form of direct and indirect trustworthiness. Bayesian network employs node's online duration, link speed, routing forwarding ratio and surplus energy as four entities to compute direct trustworthiness. Indirect trustworthiness is archived by node's single-hop neighbors' collaboration. The proposed model allows dynamic evaluation policy. Simulation shows that the model obtains dynamical, timely, precious, sensitive and fixable trustworthiness.