全局最小均方算法

本文提出了一种基于全局均方误差最小准则的全局最小均方（ＴＬＭＳ）算法，在理论上分析了它的统计特性和收敛性，用数字仿真技术优良地它的性能。     

Data decomposition using independent component analysis with prior constraints

In many data analysis problems, it is useful to consider the data as generated from a set of unknown (latent) generators or sources. The observations we make of a system are then taken to be related to these sources through some unknown function. Furthermore, the (unknown) number of underlying latent sources may be less than the number of observations. Recent developments in independent component analysis (ICA) have shown that, in the case where the unknown function linking sources to observations is linear, such data decomposition may be achieved in a mathematically elegant manner. In this paper, we extend the general ICA paradigm to include a very flexible source model, prior constraints and conditioning on sets of intermediate variables so that ICA forms one part of a hierarchical system. We show that such an approach allows for efficient discovery of hidden representation in data and for unsupervised data partitioning.     

Concealed weapon detection and visualization in a synthesized image

Images acquired by heterogeneous image sensors may provide complementary information about the scene, for instance, the visual image can provide personal identification information like the facial pattern while the infrared (IR) or millimeter wave image can detect the suspicious regions of concealed weapons. Usually, a technique, namely multiresolution pixel-level image fusion is applied to integrate the information from multi-sensor images. However, when the images are significantly different, the performance of the multiresolution fusion algorithms is not always satisfactory. In this study, a new strategy consisting of two steps is proposed. The first step is to use an unsupervised fuzzy k-means clustering to detect the concealed weapon from the IR image. The detected region is embedded in the visual image in the second step and this process is implemented with a multiresolution mosaic technique. Therefore, the synthesized image retains the quality comparable to the visual image while the region of the concealed weapon is highlighted and enhanced. The experimental results indicate the efficiency of the proposed approach.This material is based on part of the work carried out at the SPCR laboratory of Lehigh University and the work is partially supported by the U. S. Army Research Office under grant number DAAD19-00-1-0431. The content of the information does not necessarily reflect the position or the policy of the federal government, and no official endorsement should be inferred.     

基于双重权重偏差建模的无监督域适应

无监督域适应(Unsupervised Domain Adaptation,UDA)是一类新兴的机器学习范式,其通过对源域知识在无标记目标域上的迁移利用,来促进目标域模型的训练。为建模源域与目标域之间的域分布差异,最大均值差异(Maximum Mean Discrepancy,MMD)建模被广泛应用,其对UDA的性能提升起到了有效的促进作用。然而,这些方法通常忽视了领域之间对应类规模与类分布等结构信息,因为目标域与源域的数据类规模与数据分布通常并非一致。为此,文中提出了一种基于跨域类和数据样本双重加权的无监督域适应模型(Sample weighted and Class weighted based Unsupervised Domain Adaptation Network,SCUDAN)。具体而言,一方面,通过源域类层面的适应性加权来调整源域类权重,以实现源域与目标域之间的类分布对齐;另一方面,通过目标域样本层面的适应性加权来调整目标域样本权重,以实现目标域与源域类中心的对齐。此外,文中还提出了一种CEM(Classification Expectation Maximization)优化算法,以实现对SCUDAN的优化求解。最后,通过对比实验和分析,验证了所提模型和算法的有效性。     

正则化半监督判别分析方法

为了克服加权线性判别分析(WLDA)只利用有标签的训练样本而不能反映样本数据流形结构的缺点,提出一种正则化的半监督判别分析方法。首先构建所有样本的近邻图来估计数据的局部流形结构,然后将此作为正则项引入WLDA的准则函数中。该方法避免了类内散度矩阵奇异,同时保持了样本数据的判别结构和几何结构。在ORL和YALE人脸数据库上的实验结果证明了该算法的有效性。     

Unsupervised and simultaneous training of multiple object detectors from unlabeled surveillance video

Object detection is an essential component in automated vision-based surveillance systems. In general, object detectors are constructed using training examples obtained from large annotated data sets. The inevitable limitations of typical training data sets make such supervised methods unsuitable for building generic surveillance systems applicable to a wide variety of scenes and camera setups. In our previous work we proposed an unsupervised method for learning and detecting the dominant object class in a general dynamic scene observed by a static camera. In this paper, we investigate the possibilities to expand the applicability of this method to the problem of multiple dominant object classes. We propose an idea on how to approach this expansion, and perform an evaluation of this idea using two representative surveillance video sequences.     

HSP-MFL: A High-level Semantic Property driven Multi-task Feature Learning Network for unsupervised person Re-ID

Unsupervised person re-identification aims to distinguish different pedestrians from discriminative representations on the basis of unlabeled data. Currently, most unsupervised Re-ID approaches explore visual representations to generate pseudo-labels for model’s training, which may suffer from background noise and semantic loss. To tackle this problem, this paper proposes a High-level Semantic Property driven Multi-task Feature Learning Network (HSP-MFL) to firstly introduce three high-level semantic properties for unsupervised person Re-ID. Technically, we design a novel Multiple Feature Fusion Module (MFFM) to deeply explore the complex correlation among multiple semantic and visual features to capture the discriminative feature cues, as well as a multi-task training scheme to generate robust fusion features. The architecture is quite simple and does not consume extra labeling costs. Extensive experiments on three datasets demonstrate that both high-level semantic properties and multi-task learning are effective in performance improvement, yielding SOTA mAPs for unsupervised person Re-ID.     

Unsupervised learning of multi-task deep variational model

We propose a general deep variational model (reduced version, full version as well as the extension) via a comprehensive fusion approach in this paper. It is able to realize various image tasks in a completely unsupervised way without learning from samples. Technically, it can properly incorporate the CNN based deep image prior (DIP) architecture into the classic variational image processing models. The minimization problem solving strategy is transformed from iteratively minimizing the sub-problem for each variable to automatically minimizing the loss function by learning the generator network parameters. The proposed deep variational (DV) model contributes to the high order image edition and applications such as image restoration, inpainting, decomposition and texture segmentation. Experiments conducted have demonstrated significant advantages of the proposed deep variational model in comparison with several powerful techniques including variational methods and deep learning approaches.     

Unsupervised adaptive neural-fuzzy inference system for solving differential equations

There has been a growing interest in combining both neural network and fuzzy system, and as a result, neuro-fuzzy computing techniques have been evolved. ANFIS (adaptive network-based fuzzy inference system) model combined the neural network adaptive capabilities and the fuzzy logic qualitative approach. In this paper, a novel structure of unsupervised ANFIS is presented to solve differential equations. The presented solution of differential equation consists of two parts; the first part satisfies the initial/boundary condition and has no adjustable parameter whereas the second part is an ANFIS which has no effect on initial/boundary conditions and its adjustable parameters are the weights of ANFIS. The algorithm is applied to solve differential equations and the results demonstrate its accuracy and convince us to use ANFIS in solving various differential equations.