Restoration of images corrupted by mixed Gaussian-impulse noise via l1-l0 minimization

In this paper, we study the restoration of images corrupted by Gaussian plus impulse noise, and propose a l₁-l₀ minimization approach where the l₁ term is used for impulse denoising and the l₀ term is used for a sparse representation over certain unknown dictionary of images patches. The main algorithm contains three phases. The first phase is to identify the outlier candidates which are likely to be corrupted by impulse noise. The second phase is to recover the image via dictionary learning on the free-outlier pixels. Finally, an alternating minimization algorithm is employed to solve the proposed minimization energy function, leading to an enhanced restoration based on the recovered image in the second phase. Experimental results are reported to compare the existing methods and demonstrate that the proposed method is better than the other methods.     

加权局部特征结合判别式字典的目标跟踪

目的 当前大多数基于稀疏表示的跟踪方法只考虑全局特征或局部特征的最小重构误差,没有充分利用稀疏编码系数,或者忽略了字典判别性的作用,尤其当目标被相似物遮挡时,往往会导致跟踪目标丢失。针对上述问题,提出一种新的基于判别式字典和加权局部特征的稀疏外观模型(SPAM-DDWF)跟踪算法。方法 首先利用Fisher准则学习判别式字典,对提取的局部特征进行结构性分析来区分目标和背景,其次,提出一种新的基于加权的相似性度量方法来处理遮挡问题,从而提高跟踪的精确度。此外,基于重构系数的权重更新策略,使算法能更好地适应跟踪目标的外观变化,并降低了遮挡发生时跟踪漂移的概率。结果 在多个基准图像序列上,与多种流行方法对比,本文算法在光照变化、复杂背景、遮挡等场景中保持较高的跟踪成功率与较低的漂移误差。平均成功率和漂移误差分别为76.8%和3.7。结论 实验结果表明,本文算法具有较好的有效性和鲁棒性,尤其在目标被相似物遮挡的情况下,也能较准确地跟踪到目标。     

Image super-resolution by dictionary concatenation and sparse representation with approximate L0 norm minimization

This paper proposes a different image super-resolution (SR) reconstruction scheme, based on the newly advanced results of sparse representation and the recently presented SR methods via this model. Firstly, we online learn a subsidiary dictionary with the degradation estimation of the given low-resolution image, and concatenate it with main one offline learned from many natural images with high quality. This strategy can strengthen the expressive ability of dictionary atoms. Secondly, the conventional matching pursuit algorithms commonly use a fixed sparsity threshold for sparse decomposition of all image patches, which is not optimal and even introduces annoying artifacts. Alternatively, we employ the approximate L₀ norm minimization to decompose accurately the patch over its dictionary. Thus the coefficients of representation with variant number of nonzero items can exactly weight atoms for those complicated local structures of image. Experimental results show that the proposed method produces high-resolution images that are competitive or superior in quality to results generated by similar techniques.     

变分调整约束下的反向低秩稀疏学习目标跟踪

目的 低秩稀疏学习目标跟踪算法在目标快速运动和严重遮挡等情况下容易出现跟踪漂移现象,为此提出一种变分调整约束下的反向低秩稀疏学习目标跟踪算法。方法 采用核范数凸近似低秩约束描述候选粒子间的时域相关性,去除不相关粒子,适应目标外观变化。通过反向稀疏表示描述目标表观,用候选粒子稀疏表示目标模板,减少在线跟踪中L₁优化问题的数目,提高跟踪效率。在有界变差空间利用变分调整对稀疏系数差分建模,约束目标表观在相邻帧间具有较小变化,但允许连续帧间差异存在跳跃不连续性,以适应目标快速运动。结果 实验利用OTB（object tracking benchmark）数据集中的4组涵盖了严重遮挡、快速运动、光照和尺度变化等挑战因素的标准视频序列进行测试,定性和定量对比了本文算法与5种热点算法的跟踪效果。定性分析基于视频序列的主要挑战因素进行比较,定量分析通过中心点位置误差（central pixel error,CPE）比较跟踪算法的精度。与CNT（convolutional networks training）、SCM（sparse collaborative model）、IST（inverse sparse tracker）、DDL（discriminative dictionary learning）和LLR（locally low-rank representation）算法相比,平均CPE值分别提高了2.80、4.16、13.37、35.94和41.59。实验结果表明,本文算法达到了较高的跟踪精度,对上述挑战因素更具鲁棒性。结论 本文提出的跟踪算法,综合了低秩稀疏学习和变分优化调整的优势,在复杂场景下具有较高的跟踪精度,特别是对严重遮挡和快速运动情况的有效跟踪更具鲁棒性。     

Robust visual tracking using information theoretical learning

This paper presents a novel online object tracking algorithm with sparse representation for learning effective appearance models under a particle filtering framework. Compared with the state-of-the-art ? ₁ sparse tracker, which simply assumes that the image pixels are corrupted by independent Gaussian noise, our proposed method is based on information theoretical Learning and is much less sensitive to corruptions; it achieves this by assigning small weights to occluded pixels and outliers. The most appealing aspect of this approach is that it can yield robust estimations without using the trivial templates adopted by the previous sparse tracker. By using a weighted linear least squares with non-negativity constraints at each iteration, a sparse representation of the target candidate is learned; to further improve the tracking performance, target templates are dynamically updated to capture appearance changes. In our template update mechanism, the similarity between the templates and the target candidates is measured by the earth movers’ distance(EMD). Using the largest open benchmark for visual tracking, we empirically compare two ensemble methods constructed from six state-of-the-art trackers, against the individual trackers. The proposed tracking algorithm runs in real-time, and using challenging sequences performs favorably in terms of efficiency, accuracy and robustness against state-of-the-art algorithms.     

Robust visual tracking via bag of superpixels

The Bag of Words (BoW) model is one of the most popular and effective image representation methods and has been drawn increasing interest in computer vision filed. However, little attention is paid on it in visual tracking. In this paper, a visual tracking method based on Bag of Superpixels (BoS) is proposed. In BoS, the training samples are oversegmented to generate enough superpixel patches. Then K-means algorithm is performed on the collected patches to form visual words of the target and a superpixel codebook is constructed. Finally the tracking is accomplished via searching for the highest likelihood between candidates and codebooks within Bayesian inference framework. In this process, an effective updating scheme is adopted to help our tracker resist occlusions and deformations. Experimental results demonstrate that the proposed method outperforms several state-of-the-art trackers.     

判别稀疏表示鲁棒快速视觉跟踪

目的 L1跟踪对局部遮挡具有良好的鲁棒性,但存在易产生模型漂移和计算速度慢的问题。针对这两个问题,该文提出了一种基于判别稀疏表示的视觉跟踪方法。方法 考虑到背景和遮挡信息的干扰,提出了一种判别稀疏表示模型,并基于块坐标优化原理,采用学习迭代收缩阈值算法和软阈值操作设计出了表示模型的快速求解算法。结果 在8组图像序列中,该文方法与现有的4种经典跟踪方法分别在鲁棒性和稀疏表示的计算时间方面进行了比较。在鲁棒性的定性和定量比较实验中,该文方法不仅表现出了对跟踪过程中的多种干扰因素具有良好的适应能力,而且在位置误差阈值从0~50像素的变化过程中,其精度曲线均优于实验中的其他方法;在稀疏表示的计算时间方面,在采用大小为16×16和32×32的模板进行跟踪时,该文算法的时间消耗分别为0.152 s和0.257 s,其时效性明显优于实验中的其他方法。结论 与经典的跟踪方法相比,该文方法能够在克服遮挡、背景干扰和外观改变等诸多不良因素的同时,实现快速目标跟踪。由于该文方法不仅具有较优的稀疏表示计算速度,而且能够克服多种影响跟踪鲁棒性的干扰因素,因此可以将其应用于视频监控和体育竞技等实际场合。     

协同结构稀疏重构的判别性视觉跟踪

基于稀疏表示的表观似然模型在目标跟踪领域具有广泛的应用,但是这种单一产生式目标表观模型并未考虑完整的判别性结构信息,容易受复杂背景的干扰。为了缓解由该问题造成的目标跟踪漂移,提出了一种目标表观字典和背景字典协同结构稀疏重构优化的视觉跟踪方法。通过构建一个有判别力的基于稀疏表示的表观似然模型,实现了对目标表观模型更为准确的描述。通过合理选择约束候选目标区域和候选背景区域的稀疏系数,在表观似然模型中引入判别式信息,以进一步揭示候选目标区域的潜在相关性和候选背景区域的结构关系,从而更加准确地学习候选目标区域的表观模型。大量有挑战性的视频序列上的实验结果验证了算法在复杂背景下跟踪的鲁棒性,与其他相关算法的对比实验也体现了该算法的优越性。     

Evolution-enhanced multiscale overcomplete dictionaries learning for image denoising

In this paper, a multiscale overcomplete dictionary learning approach is proposed for image denoising by exploiting the multiscale property and sparse representation of images. The images are firstly sparsely represented by a translation invariant dictionary and then the coefficients are denoised using some learned multiscale dictionaries. Dictionaries learning can be reduced to a non-convex l₀-norm minimization problem with multiple variables, so an evolution-enhanced algorithm is proposed to alternately optimize the variables. Some experiments are taken on comparing the performance of our proposed method with its counterparts on some benchmark natural images, and the superiorities of our proposed method to its counterparts can be observed in both the visual result and some numerical guidelines.     

Object tracking via collaborative multi-task learning and appearance model updating

In this paper, we propose a novel visual tracking algorithm using the collaboration of generative and discriminative trackers under the particle filter framework. Each particle denotes a single task, and we encode all the tasks simultaneously in a structured multi-task learning manner. Then, we implement generative and discriminative trackers, respectively. The discriminative tracker considers the overall information of object to represent the object appearance; while the generative tracker takes the local information of object into account for handling partial occlusions. Therefore, two models are complementary during the tracking. Furthermore, we design an effective dictionary updating mechanism. The dictionary is composed of fixed and variational parts. The variational parts are progressively updated using Metropolis–Hastings strategy. Experiments on different challenging video sequences demonstrate that the proposed tracker performs favorably against several state-of-the-art trackers.     

判别稀疏表示与在线字典学习的运动目标跟踪

针对传统稀疏表示不能有效区分目标和背景的缺点,提出一种判别稀疏表示算法,这种算法在传统稀疏表示目标函数中加入一个判别函数,大大降低干扰因素对目标跟踪的影响。基于判别稀疏表示和[?1]约束,提出一种在线字典学习算法升级目标模板,有效降低背景信息对目标模板的影响。提取目标梯度方向的直方图（HOG）特征,利用其对光照和形变等复杂环境具有较强鲁棒性的优点,实现对目标更稳定的跟踪。实验结果表明,与现有跟踪方法相比,该算法的跟踪效果更好。     

Effective texture classification by texton encoding induced statistical features

Effective and efficient texture feature extraction and classification is an important problem in image understanding and recognition. Recently, texton learning based texture classification approaches have been widely studied, where the textons are usually learned via K-means clustering or sparse coding methods. However, the K-means clustering is too coarse to characterize the complex feature space of textures, while sparse texton learning/encoding is time-consuming due to the l₀-norm or l₁-norm minimization. Moreover, these methods mostly compute the texton histogram as the statistical features for classification, which may not be effective enough. This paper presents an effective and efficient texton learning and encoding scheme for texture classification. First, a regularized least square based texton learning method is developed to learn the dictionary of textons class by class. Second, a fast two-step l₂-norm texton encoding method is proposed to code the input texture feature over the concatenated dictionary of all classes. Third, two types of histogram features are defined and computed from the texton encoding outputs: coding coefficients and coding residuals. Finally, the two histogram features are combined for classification via a nearest subspace classifier. Experimental results on the CUReT, KTH_TIPS and UIUC datasets demonstrated that the proposed method is very promising, especially when the number of available training samples is limited.     

I-VITAL: Information aided visual tracking with adversarial learning

With the advent of convolutional neural networks (CNN), MDNet and the Siamese trackers posed tracking as supervised learning. They model an object’s presence using classification (foreground and background) and location using regression. For the first time, we have brought probability distribution into the CNN framework for tracking. We have selected “Information maximization Generative Adversarial Network (InfoGAN)” to couple the target and background classes with two unique Gaussian distributions. This paper highlights the use of InfoGAN in information extraction & feedback to improve the tracking framework. Specifically, the novel features proposed in this tracking framework are (i) Coupling of unique probability distributions to target and background classes and (ii) Unsupervised tracker status (success/ failure) identification and correction through information feedback. We demonstrated the efficacy of the proposed I-VITAL tracker in visual tracking with experimental comparisons on well-known data sets such as GOT10K, VOT2020, and OTB-2015. Compared with base works, the proposed tracker has improved performance in locating the object of interest.     

Robust object tracking via multi-scale patch based sparse coding histogram

There are many visual tracking algorithms that are based on sparse representation appearance model. Most of them are modeled by local patches with fixed patch scale, which make trackers less effective when objects undergone appearance changes such as illumination variation, pose change or partial occlusion. To solve the problem, a novel appearance representation model is proposed via multi-scale patch based sparse coding histogram for robust visual tracking. In this paper, the appearance of an object is modeled by different scale patches, which are represented by sparse coding histogram with different scale dictionaries. Then a similarity measure is applied to the calculation of the distance between the sparse coding histograms of target candidate and target template. Finally, the similarity score of the target candidate is passed to a particle filter to estimate the target state sequentially in the tracking process. Additionally, in order to decrease the visual drift caused by partial occlusion, an occlusion handling strategy is adopted, which takes the spatial information of multi-scale patches and occlusion into account. Based on the experimental results on some benchmarks of video sequences, our tracker outperforms state-of-the-art tracking methods.     

A parallel and robust object tracking approach synthesizing adaptive Bayesian learning and improved incremental subspace learning

This paper presents a novel tracking algorithm which integrates two complementary trackers. Firstly, an improved Bayesian tracker(B-tracker) with adaptive learning rate is presented. The classification score of B-tracker reflects tracking reliability, and a low score usually results from large appearance change. Therefore, if the score is low, we decrease the learning rate to update the classifier fast so that B-tracker can adapt to the variation and vice versa. In this way, B-tracker is more suitable than its traditional version to solve appearance change problem. Secondly, we present an improved incremental subspace learning method tracker(Stracker). We propose to calculate projected coordinates using maximum posterior probability, which results in a more accurate reconstruction error than traditional subspace learning tracker. Instead of updating at every time, we present a stopstrategy to deal with occlusion problem. Finally, we present an integrated framework(BAST), in which the pair of trackers run in parallel and return two candidate target states separately. For each candidate state, we define a tracking reliability metrics to measure whether the candidate state is reliable or not, and the reliable candidate state will be chosen as the target state at the end of each frame. Experimental results on challenging sequences show that the proposed approach is very robust and effective in comparison to the state-of-the-art trackers.     

红外可见光目标的空间直方图表示与联合跟踪

目的 针对融合跟踪中的实时性和准确性问题,提出一种基于二阶空间直方图联合的红外与可见光目标融合跟踪算法。方法 该算法以二阶空间直方图为目标表示模型,通过将红外的目标相似度和可见光的目标相似度进行加权融合,来构建新的目标函数;并依据核跟踪推理机制导出目标的联动位移公式;最后使用均值漂移程序实现目标的自动搜索。此外,算法还实现了融合权值的自适应调节和目标模型的在线更新。结果 实验中选取了4组典型的红外可见光视频对进行跟踪,测试了算法在夜间环境、背景阴影、目标交汇与拥簇,以及目标遮挡等场合下的跟踪性能,并与L₁跟踪器（L1T）、基于区域模糊动态融合的跟踪器（FRD）,以及基于联合直方图的跟踪器在平均中心误差、平均重叠率、成功率以及平均跟踪时间等指标上进行了定量比较,得到各算法在这4组视频上的对应性能指标数据分别为本文算法（6.664,0.702,0.921,0.009）、L1T跟踪红外目标（25.53,0.583,0.742,0.363）、L1T跟踪可见光目标（31.21,0.359,0.459,0.293）、FRD（10.73,0.567,0.702,0.565）、JHT（15.07,0.622,0.821,0.001）,发现本文算法的平均准确率比其他跟踪算法分别高约23%、14%和8%,而平均成功率分别高约32%、46%和10%。结论 本文算法在处理场景拥簇、光照变化以及空间信息保持等方面要优于传统的单源跟踪方法,适用于夜间环境、背景阴影以及背景拥簇等场景下目标跟踪,对帧频为30 帧/s的视频数据,算法可同时在线跟踪到4个目标。     

结合双模板融合与孪生网络的鲁棒视觉目标跟踪

目的 视觉目标跟踪算法主要包括基于相关滤波和基于孪生网络两大类。前者虽然精度较高但运行速度较慢,无法满足实时要求。后者在速度和精度方面取得了出色的跟踪性能,然而,绝大多数基于孪生网络的目标跟踪算法仍然使用单一固定的模板,导致算法难以有效处理目标遮挡、外观变化和相似干扰物等情形。针对当前孪生网络跟踪算法的不足,提出了一种高效、鲁棒的双模板融合目标跟踪方法(siamese tracker with double template fusion,Siam-DTF)。方法 使用第1帧的标注框作为初始模板,然后通过外观模板分支借助外观模板搜索模块在跟踪过程中为目标获取合适、高质量的外观模板,最后通过双模板融合模块,进行响应图融合和特征融合。融合模块结合了初始模板和外观模板各自的优点,提升了算法的鲁棒性。结果 实验在3个主流的目标跟踪公开数据集上与最新的9种方法进行比较,在OTB2015(object tracking benchmark 2015)数据集中,本文方法的AUC(area under curve)得分和精准度分别为0.701和0.918,相比于性能第2的SiamRPN++(siamese region proposal network++)算法分别提高了0.6%和1.3%;在VOT2016(visual object tracking 2016)数据集中,本文方法取得了最高的期望平均重叠(expected average overlap,EAO)和最少的失败次数,分别为0.477和0.172,而且EAO得分比基准算法SiamRPN++提高了1.6%,比性能第2的SiamMask_E算法提高了1.1%;在VOT2018数据集中,本文方法的期望平均重叠和精确度分别为0.403和0.608,在所有算法中分别排在第2位和第1位。本文方法的平均运行速度达到47帧/s,显著超出跟踪问题实时性标准要求。结论 本文提出的双模板融合目标跟踪方法有效克服了当前基于孪生网络的目标跟踪算法的不足,在保证算法速度的同时有效提高了跟踪的精确度和鲁棒性,适用于工程部署与应用。     

正例投票下的L1目标跟踪算法

目的 传统的L1稀疏表示目标跟踪,是将所有候选目标表示为字典模板的线性组合,只考虑了字典模板的整体信息,没有分析目标的局部结构。针对该方法在背景杂乱时容易出现跟踪漂移的问题,提出一种基于正例投票的目标跟踪算法。方法 本文将目标表示成图像块粒子的组合,考虑目标的局部结构。在粒子滤波框架内,构建图像块粒子置信函数和相似性函数,提取正例图像块。最终通过正例权重投票估计跟踪目标的最佳位置。结果 在14组公测视频序列上进行跟踪实验,与多种优秀的目标跟踪算法相比,本文跟踪算法在目标受到背景杂乱、遮挡、光照变化等复杂环境干扰下最为稳定,重叠率达到了0.7,且取得了最低的平均跟踪误差5.90,反映了本文算法的可靠性和有效性。结论 本文正例投票下的L1目标跟踪算法,与经典方法相比,能够解决遮挡、光照变化和快速运动等问题的同时,稳定可靠地实现背景杂乱序列的鲁棒跟踪。     

基于双准则自适应融合的目标跟踪算法

针对单一评判准则较难适应复杂环境下的目标跟踪问题,提出了一种基于双评判准则自适应融合的跟踪算法。在该算法中,空间直方图被用作目标表示模型,候选目标与目标模板之间的相似度、以及候选目标与其邻近背景区域之间的对比度被作为目标评判双准则,而目标函数(或似然函数)则由两个准则的加权融合而成。算法是在粒子滤波框架下实现的目标搜索,并采用了模糊逻辑对相似度和对比度的权值进行自适应调节。对人、动物等多个挑战性运动目标的跟踪结果表明,与增量学习跟踪、ι₁跟踪等最新跟踪器相比,所提算法在处理目标的遮挡、形变、旋转以及表观变化方面的综合性能更好,其成功率和平均重叠率指标分别在80%和0.76以上。     

快速深度学习的鲁棒视觉跟踪

目的 基于深度学习的视觉跟踪算法具有跟踪精度高、适应性强的特点,但是,由于其模型参数多、调参复杂,使得算法的时间复杂度过高。为了提升算法的效率,通过构建新的网络结构、降低模型冗余,提出一种快速深度学习的算法。方法 鲁棒特征的提取是视觉跟踪成功的关键。基于深度学习理论,利用海量数据离线训练深度神经网络,分层提取描述图像的特征;针对网络训练时间复杂度高的问题,通过缩小网络规模得以大幅缓解,实现了在GPU驱动下的快速深度学习;在粒子滤波框架下,结合基于支持向量机的打分器的设计,完成对目标的在线跟踪。结果 该方法精简了特征提取网络的结构,降低了模型复杂度,与其他基于深度学习的算法相比,具有较高的时效性。系统的跟踪帧率总体保持在22帧/s左右。结论 实验结果表明,在目标发生平移、旋转和尺度变化,或存在光照、遮挡和复杂背景干扰时,本文算法能够实现比较稳定和相对快速的目标跟踪。但是,对目标的快速移动和运动模糊的鲁棒性不够高,容易受到相似物体的干扰。