Complete Model Research Articles

Lightweight Super-Resolution Model for Complete Model Copyright Protection

Lightweight Super-Resolution Model for Complete Model Copyright Protection

A Cortical-Inspired Contour Completion Model Based on Contour Orientation and Thickness.

An extended four-dimensional version of the traditional Petitot-Citti-Sarti model on contour completion in the visual cortex is examined. The neural configuration space is considered as the group of similarity transformations, denoted as M=SIM(2). The left-invariant subbundle of the tangent bundle models possible directions for establishing neural communication. The sub-Riemannian distance is proportional to the energy expended in interneuron activation between two excited border neurons. According to the model, the damaged image contours are restored via sub-Riemannian geodesics in the space M of positions, orientations and thicknesses (scales). We study the geodesic problem in M using geometric control theory techniques. We prove the existence of a minimal geodesic between arbitrary specified boundary conditions. We apply the Pontryagin maximum principle and derive the geodesic equations. In the special cases, we find explicit solutions. In the general case, we provide a qualitative analysis. Finally, we support our model with a simulation of the association field.

GFedKG: GNN-based federated embedding model for knowledge graph completion

Knowledge Graph (KG) completion through reasoning over KGs is one of the most concerned issue in research around knowledge graphs. Current work focuses more on studying individual KG, while a series of related multi-source KGs are more common in real applications. In order to effectively exert their complementary roles in completing each KG under the premise of ensuring data privacy, federated learning becomes the focus of attention. However, existing methods ignore the explicit complementarity among KGs in terms of structural information. And they generally align embeddings of overlapping instances across clients, which constrains flexibility of each client model. Based on this, we propose the first GNN-based federated embedding framework for multiple knowledge graph completion, GFedKG. It focuses on the complementary role among KGs in topology structures and adaptively generates different aggregated neighboring features for entities in distinct clients. Such treatment achieves some consistency among clients while retaining the flexibility of each client model to characterize respective KG. Moreover, entity embeddings that represent the features of entities are better privacy-preserved since neighboring embeddings are transmitted instead of them. And with the aim of realizing more refined and accurate aggregation of neighboring features from multiple clients, we innovatively propose a context-aware graph neural network, along with the context-enhanced evaluation function. Homophily neighbors are also taken into account to cope with the low-homophily aspect of KGs. Extensive experiments demonstrate the advantages of GFedKG.

Cyclic tensor singular value decomposition with applications in low-rank high-order tensor recovery

The rapid advancements in emerging technologies have increased the demand for recovery tasks involving high-dimensional data with complex structures. Effectively utilizing tensor decomposition techniques to capture the low-rank structure of such data is crucial. Recently, the high-order t-SVD has demonstrated strong adaptability. However, this decomposition approach can only capture the low-rank correlation of two modes along other modes individually, while disregarding the structural correlation between different modes. In this paper, we propose a novel cyclic tensor singular value decomposition (CTSVD) method that effectively characterizes the low-rank structures of high-order tensors along all modes. Specifically, our method decomposes an order-N tensor into N factor tensors and one core tensor, connecting them using a defined mode-k tensor-tensor product (t-product). Building upon this, we establish the corresponding tensor rank and its convex relaxation. To address the issue of dimensional imbalance between adjacent modes in high-dimensional data, we propose and integrate a square reshaping strategy into the recovery models for tensor completion (TC) and tensor principal component analysis (TRPCA) tasks. Effective alternating direction method of multipliers (ADMM)-based algorithms are designed to these tasks. Extensive experiments on both synthetic and real data demonstrate that our methods outperform state-of-the-art approaches.

A nonlinear high-order transformations-based method for high-order tensor completion

The high-order tensor nuclear norm model (HTNN) has recently shown promising results in tensor completion problems. The HTNN-based approaches rely on the low-rank structure of tensor slices during reversible transformations. However, under reversible transformations, the low-rank configuration within the slice-wise modality of the tensor’s structure is not markedly pronounced. In order to more effectively describe the low-rank characteristics, we propose a model based on the nonlinear high-order transform-based tensor nuclear norm (NHTNN). Specifically, our framework consists of a linearly semi-orthogonal transformation along the high-dimensional modality and an element-wise nonlinear transformation. We introduce a model for tensor completion, grounded in the suggested measure of tensor low-rank, i.e., NHTNN. Utilizing this non-convex nonlinear model, we formulate a proximal alternating minimization (PAM) algorithm, establishing its convergence through a rigorous proof. In experiments on datasets such as hyperspectral videos (HSVs) and color videos (CVs), our approach demonstrates superior quantitative numerical results and qualitative visual effects compared to cutting-edge tensor completion techniques.

An experimentally validated approach to automated biological evidence generation in drug discovery using knowledge graphs

Explaining predictions for drug repositioning with biological knowledge graphs is a challenging problem. Graph completion methods using symbolic reasoning predict drug treatments and associated rules to generate evidence representing the therapeutic basis of the drug. Yet the vast amounts of generated paths that are biologically irrelevant or not mechanistically meaningful within the context of disease biology can limit utility. We use a reinforcement learning based knowledge graph completion model combined with an automatic filtering approach that produces the most relevant rules and biological paths explaining the predicted drug’s therapeutic connection to the disease. In this work we validate the approach against preclinical experimental data for Fragile X syndrome demonstrating strong correlation between automatically extracted paths and experimentally derived transcriptional changes of selected genes and pathways of drug predictions Sulindac and Ibudilast. Additionally, we show it reduces the number of generated paths in two case studies, 85% for Cystic fibrosis and 95% for Parkinson’s disease.

A Sparsity-Invariant Model via Unifying Depth Prediction and Completion

The development of a sparse-invariant depth completion model capable of handling varying levels of input depth sparsity is highly desirable in real-world applications. However, existing sparse-invariant models tend to degrade when the input depth points are extremely sparse. In this paper, we propose a new model that combines the advantageous designs of depth completion and monocular depth estimation tasks to achieve sparse invariance. Specifically, we construct a dual-branch architecture with one branch dedicated to depth prediction and the other to depth completion. Additionally, we integrate the multi-scale local planar module in the decoders of both branches. Experimental results on the NYU Depth V2 benchmark and the OPPO prototype dataset equipped with the Spot-iToF316 sensor demonstrate that our model achieves reliable results even in cases with irregularly distributed, limited or absent depth information.

Tensor nonconvex unified prior for tensor recovery

Tensor data, such as hyperspectral images and multi-frame videos, have gained significant attention in practical applications. However, the inherent degradation phenomena during data acquisition, including noise and missing pixels, give rise to a series of ill-posed inverse problems that need to be addressed. Currently, the rational exploration of prior knowledge for tensor recovery, including global low-rankness and local smoothness, has emerged as a common concern. Inspired by recent notable works, this paper proposes a novel tensor non-convex unified prior term, which employs weighted tensor Schatten p-norm as a rank surrogate function in the gradient domain. The new prior can yield a regularizer that effectively captures low-rankness and smoothness, and is applied to tensor completion and tensor robust principal component analysis models. An efficient algorithm is developed by using the alternating direction method of multipliers and its convergence analysis is also provided. Extensive experimental results demonstrate that the proposed method outperforms the state-of-the-art methods, particularly in cases of high missing rates and strong noise levels.

CoNoT: Coupled Nonlinear Transform-Based Low-Rank Tensor Representation for Multidimensional Image Completion.

Recently, the transform-based tensor nuclear norm (TNN) methods have shown promising performance and drawn increasing attention in tensor completion (TC) problems. The main idea of these methods is to exploit the low-rank structure of frontal slices of the tensor under the transform. However, the transforms in TNN methods usually treat all modes equally and do not consider the different traits of different modes (i.e., spatial and spectral/temporal modes). To address this problem, we suggest a new low-rank tensor representation based on the coupled nonlinear transform (called CoNoT) for a better low-rank approximation. Concretely, spatial and spectral/temporal transforms in the CoNoT, respectively, exploit the different traits of different modes and are coupled together to boost the implicit low-rank structure. Here, we use the convolutional neural network (CNN) as the CoNoT, which can be learned solely from an observed multidimensional image in an unsupervised manner. Based on this low-rank tensor representation, we build a new multidimensional image completion model. Moreover, we also propose an enhanced version (called Ms-CoNoT) to further exploit the spatial multiscale nature of real-world data. Extensive experiments on real-world data substantiate the superiority of the proposed models against many state-of-the-art methods both qualitatively and quantitatively.

Fine-Grained Entity-Type Completion Based on Neighborhood-Attention and Cartesian–Polar Coordinates Mapping

Entities refer to things that exist objectively, and entity types are concepts abstracted from entities that have the same features or properties. However, the entity types in the knowledge graph are always incomplete. Currently, the main approach for predicting missing entity types is to learn structured representations of entities and types separately, which ignores neighborhood semantic knowledge of the entity. Therefore, this paper proposes the aggregation neighborhood semantics model for type completion (ANSTC), which extracts neighborhood triple features of target entities with two attentional mechanisms. Meanwhile, the spatial mapping module in ANSTC maps entities from Cartesian coordinate to Polar coordinate system, which can map similar vectors onto a concentric circle and then rotate the angle according to the fine-grained difference to achieve entity-to-type transformation. Moreover, we add semantic features from text to the entity representations to enrich semantics. Through experimental comparison on the FB15K and YAGO43K dataset, we get similar results to the baseline. We also construct person dataset in computer domain, and the values of MRR, Hit@1, Hit@3 and Hit@10 are improved compared with the ConnectE model. The experimental results demonstrate that our model can effectively predict the fine-grained entity types in the domain dataset, and achieve state-of-the-art performance.

Non-Autoregressive Line-Level Code Completion

Software developers frequently use code completion tools to accelerate software development by suggesting the following code elements. Researchers usually employ AutoRegressive (AR) decoders to complete code sequences in a left-to-right, token-by-token fashion. To improve the accuracy and efficiency of code completion, we argue that tokens within a code statement have the potential to be predicted concurrently. In this article, we first conduct an empirical study to analyze the dependency among the target tokens in line-level code completion. The results suggest that it is potentially practical to generate all statement tokens in parallel. To this end, we introduce SANAR, a simple and effective syntax-aware non-autoregressive model for line-level code completion. To further improve the quality of the generated code, we propose an adaptive and syntax-aware sampling strategy to boost the model’s performance. The experimental results obtained from two widely used datasets indicate that our model outperforms state-of-the-art code completion approaches of similar model size by a considerable margin, and is faster than these models with up to 9× speed-up. Moreover, the extensive results additionally demonstrate that the enhancements achieved by SANAR become even more pronounced with larger model sizes, highlighting their significance.

“Sparse + Low-Rank” tensor completion approach for recovering images and videos

Recovering color images and videos from highly undersampled data is a fundamental and challenging task in face recognition and computer vision. By the multi-dimensional nature of color images and videos, in this paper, we propose a novel tensor completion approach, which is able to efficiently explore the sparsity of tensor data under the discrete cosine transform (DCT). Specifically, we introduce two “sparse + low-rank” tensor completion models as well as two implementable algorithms for finding their solutions. The first one is a DCT-based sparse plus weighted nuclear norm induced low-rank minimization model. The second one is a DCT-based sparse plus p-shrinking mapping induced low-rank optimization model. Moreover, we accordingly propose two implementable augmented Lagrangian-based algorithms for solving the underlying optimization models. A series of numerical experiments including color image inpainting and video data recovery demonstrate that our proposed approach performs better than many existing state-of-the-art tensor completion methods, especially for the case when the ratio of missing data is high.

ShallowBKGC: a BERT-enhanced shallow neural network model for knowledge graph completion

Knowledge graph completion aims to predict missing relations between entities in a knowledge graph. One of the effective ways for knowledge graph completion is knowledge graph embedding. However, existing embedding methods usually focus on developing deeper and more complex neural networks, or leveraging additional information, which inevitably increases computational complexity and is unfriendly to real-time applications. In this article, we propose an effective BERT-enhanced shallow neural network model for knowledge graph completion named ShallowBKGC. Specifically, given an entity pair, we first apply the pre-trained language model BERT to extract text features of head and tail entities. At the same time, we use the embedding layer to extract structure features of head and tail entities. Then the text and structure features are integrated into one entity-pair representation via average operation followed by a non-linear transformation. Finally, based on the entity-pair representation, we calculate probability of each relation through multi-label modeling to predict relations for the given entity pair. Experimental results on three benchmark datasets show that our model achieves a superior performance in comparison with baseline methods. The source code of this article can be obtained from https://github.com/Joni-gogogo/ShallowBKGC.

State-Building through Public Land Disposal? An Application of Matrix Completion for Counterfactual Prediction

This article examines how homestead policies, which opened vast frontier lands for settlement, influenced the development of American frontier states. It uses a treatment propensity-weighted matrix completion model to estimate the counterfactual size of these states without homesteading. In simulation studies, the method shows lower bias and variance than other estimators, particularly in higher complexity scenarios. The empirical analysis reveals that homestead policies significantly and persistently reduced state government expenditure and revenue. These findings align with continuous difference-in-differences estimates using 1.46 million land patent records. This study’s extension of the matrix completion method to include propensity score weighting for causal effect estimation in panel data, especially in staggered treatment contexts, enhances policy evaluation by improving the precision of long-term policy impact assessments.

SparseDC: Depth completion from sparse and non-uniform inputs

We propose SparseDC, a model for Depth Completion from Sparse and non-uniform inputs. Unlike previous methods focusing on completing fixed distributions on benchmark datasets (e.g., NYU with 500 points, KITTI with 64 lines), SparseDC is specifically designed to handle depth maps with poor quality in real usage. Our SparseDC makes two major contributions. First, we design a simple strategy, called SFFM, to improve the robustness under sparse inputs by explicitly filling the unstable depth features with stable image features. Second, we suggest utilizing a two-branch feature embedder in order to forecast the exact local geometry of regions with available depth values and accurate structures in regions with no depth. The key of the embedder is an uncertainty-based feature fusion module called UFFM to balance the local and long-term information extracted by CNNs and ViTs. Numerous experiments conducted both indoors and outdoors show how robust and effective our framework is when facing sparse and non-uniform input depths. And SparseDC outperforms the current state-of-the-art (SOTA) method, CFormer, with varying inputs. We use less parameters (38.2M < 45M) and achieve a +18.6% increase in the REL metric and a +9.9% increase in the RMSE metric on NYU Depth dataset. The pre-trained model and the complete code repository can be accessed at https://github.com/WHU-USI3DV/SparseDC.

MODELS OF INTERCULTURAL COMPETENCE: COMPARATIVE ANALYSIS

MODELS OF INTERCULTURAL COMPETENCE: COMPARATIVE ANALYSIS

Fusion of low-rankness and smoothness under learnable nonlinear transformation for tensor completion

The recently proposed tensor correlated total variation (t-CTV) has achieved success in tensor completion. It utilizes the low-rank structure of the gradient tensor under a unified linear transform to jointly encode low-rankness and smoothness priors. However, fixed linear transforms have inherent limitations in fully characterizing gradient tensors in different directions and adapting them to tensors from diverse categories. In this work, we propose the nonlinear tensor correlated total variation (NTCTV) regularization term that leverages the low-rank correlations of the gradient tensor under the learnable nonlinear transformation, providing a more natural approach to fuse the low-rankness and smoothness priors. Specifically, our approach learns the optimal nonlinear implicit low-rank structure of the gradient tensor along different modes separately, and then achieves the expression of fused prior information in a coupled manner. Furthermore, we propose the NTCTV-based tensor completion model and design the proximal alternating minimization (PAM) algorithm to efficiently solve the optimization model. Moreover, we provide a theoretical proof of the global convergence of the algorithm to a critical point. Comprehensive experimental results for hyperspectral images, medical images, multispectral images, and videos demonstrate that the proposed method achieves substantial quantitative and qualitative improvements over many state-of-the-art tensor completion techniques.

Multi-Channel Audio Completion Algorithm Based on Tensor Nuclear Norm

Multi-channel audio signals provide a better auditory sensation to the audience. However, missing data may occur in the collection, transmission, compression, or other processes of audio signals, resulting in audio quality degradation and affecting the auditory experience. As a result, the completeness of the audio signal has become a popular research topic in the field of signal processing. In this paper, the tensor nuclear norm is introduced into the audio signal completion algorithm, and the multi-channel audio signals with missing data are restored by using the completion algorithm based on the tensor nuclear norm. First of all, the multi-channel audio signals are preprocessed and are then transformed from the time domain to the frequency domain. Afterwards, the multi-channel audio with missing data is modeled to construct a third-order multi-channel audio tensor. In the next part, the tensor completion algorithm is used to complete the third-order tensor. The optimal solution of the convex optimization model of the tensor completion is obtained by using the convex relaxation technique and, ultimately, the data recovery of the multi-channel audio with data loss is accomplished. The experimental results of the tensor completion algorithm and the traditional matrix completion algorithm are compared using both objective and subjective indicators. The final result shows that the high-order tensor completion algorithm has a better completion ability and can restore the audio signal better.

Transforms-Based Bayesian Tensor Completion Method for Network Traffic Measurement Data Recovery

Network traffic measurement is regarded as the bedrock of next-generation network systems. Its purpose is to monitor the network traffic and provide data support for traffic engineering. For this reason, monitoring traffic data from a network-wide perspective is particularly important. However, the proliferation of network services has led to the explosive growth of network traffic, which has brought significant challenges to the measure of network-wide traffic. Therefore, how to infer network-wide traffic from partial traffic data is extremely important. In this article, a transforms-based Bayesian tensor completion (TBTC) method is proposed to infer network traffic data. First, the heterogeneous network traffic data with missing entries are organized into observation tensors according to temporal dimensions and other attributes. Second, the sparse hierarchical prior is used to induce lateral slices sparsity of factor tensors, which makes the tubal rank of the observation tensor can be estimated. Further, a variational Bayesian inference method is developed for model learning, and an efficient updating method is presented. Finally, two cases of the linear transforms-based tensor completion model are implemented in the experiments. Experimental results on two real-world network traffic datasets validate that the proposed method can efficiently and accurately recover network traffic data.

Quaternion matrix completion using untrained quaternion convolutional neural network for color image inpainting

The use of quaternions as a novel tool for color image representation has yielded impressive results in color image processing. By considering the color image as a unified entity rather than separate color space components, quaternions can effectively exploit the strong correlation among the RGB channels, leading to enhanced performance. Especially, color image inpainting tasks are highly beneficial from the application of quaternion matrix completion techniques, in recent years. However, existing quaternion matrix completion methods suffer from two major drawbacks. First, it can be difficult to choose a regularizer that captures the common characteristics of natural images, and sometimes the regularizer that is chosen based on empirical evidence may not be the optimal or efficient option. Second, the optimization process of quaternion matrix completion models is quite challenging because of the non-commutativity of quaternion multiplication. To address the two drawbacks of the existing quaternion matrix completion approaches mentioned above, this paper tends to use an untrained quaternion convolutional neural network (QCNN) to directly generate the completed quaternion matrix. This approach replaces the explicit regularization term in the quaternion matrix completion model with an implicit prior that is learned by the QCNN. Extensive quantitative and qualitative evaluations demonstrate the superiority of the proposed method for color image inpainting compared with some existing quaternion-based and tensor-based methods.