Search

The vanilla Few-shot Learning (FSL) learns to build a classifier for a new concept from one or very few target examples, with the general assumption that source and target classes are sampled from the same domain. Recently, the task of Cross-Domain Few-Shot Learning (CD-FSL) aims at tackling the FSL where there is a huge domain shift between the source and target datasets. Extensive efforts on CD-FSL have been made via either directly extending the meta-learning paradigm of vanilla FSL methods, or employing massive unlabeled target data to help learn models. In this paper, we notice that in the CD-FSL task, the few labeled target images have never been explicitly leveraged to inform the model in the training stage. However, such a labeled target example set is very important to bridge the huge domain gap. Critically, this paper advocates a more practical training scenario for CD-FSL. And our key insight is to utilize a few labeled target data to guide the learning of the CD-FSL model. Technically, we propose a novel Generalized Meta-learning based Feature-Disentangled Mixup network, namely GMeta-FDMixup. We make three key contributions of utilizing GMeta-FDMixup to address CD-FSL. Firstly, we present two mixup modules – mixup-P and mixup-M that help facilitate utilizing the unbalanced and disjoint source and target datasets. These two novel modules enable diverse image generation for training the model on the source domain. Secondly, to narrow the domain gap explicitly, we contribute a novel feature disentanglement module that learns to decouple the domain-irrelevant and domain-specific features. By stripping the domain-specific features, we alleviate the negative effects caused by the domain inductive bias. Finally, we repurpose a new contrastive learning module, dubbed ConL. ConL prevents the model from only capturing category-related features via introducing contrastive loss. Thus, the generalization ability on novel categories is improved. Extensive experimental results on two benchmarks show the superiority of our setting and the effectiveness of our method. Code and models will be released. [ABSTRACT FROM AUTHOR]

Extracting robust and discriminative local features from images plays a vital role for long term visual localization, whose challenges are mainly caused by the severe appearance differences between matching images due to the day-night illuminations, seasonal changes, and human activities. Existing solutions resort to jointly learning both keypoints and their descriptors in an end-to-end manner, leveraged on large number of annotations of point correspondence which are harvested from the structure from motion and depth estimation algorithms. While these methods show improved performance over non-deep methods or those two-stage deep methods, i.e., detection and then description, they are still struggled to conquer the problems encountered in long term visual localization. Since the intrinsic semantics are invariant to the local appearance changes, this paper proposes to learn semantic-aware local features in order to improve robustness of local feature matching for long term localization. Based on a state of the art CNN architecture for local feature learning, i.e., ASLFeat, this paper leverages on the semantic information from an off-the-shelf semantic segmentation network to learn semantic-aware feature maps. The learned correspondence-aware feature descriptors and semantic features are then merged to form the final feature descriptors, for which the improved feature matching ability has been observed in experiments. In addition, the learned semantics embedded in the features can be further used to filter out noisy keypoints, leading to additional accuracy improvement and faster matching speed. Experiments on two popular long term visual localization benchmarks (Aachen Day and Night v1.1, Robotcar Seasons) and one challenging indoor benchmark (InLoc) demonstrate encouraging improvements of the localization accuracy over its counterpart and other competitive methods. [ABSTRACT FROM AUTHOR]

Multi-object tracking (MOT), which is crucial for computer vision and video processing, has immense potential for improvement. Traditional tracking-by-detection approaches include feature-based object re-identification methods that use trained features, but these methods suffer from a lack of suitable training data. In training datasets used for MOT, every object in a video sequence must have its own location and ID. However, assigning IDs to each object in every sequence is considerably labor-intensive, and hence current MOT datasets are unsuitable for training re-identification networks. To resolve this issue, this paper proposes a novel self-supervised learning method using several short videos that contain no human-added labels, based on the idea that each video is a set of temporally corresponding image frames. We then describe how to improve tracking performance using a re-identification network trained in a self-supervised manner. In addition, ablation studies were conducted in order to define the optimal parameters, such as number of clips, data augmentation, and appropriate matching algorithms. The proposed approach achieved competitive performance compared with current best-practice methods including supervised methods, achieving MOT accuracy = 62.0% and ID F1-score = 62.7% on the MOT17 benchmark. [ABSTRACT FROM AUTHOR]

Vehicle Re-Identification (ReID) is of great significance for public security and intelligent transportation. Large and comprehensive datasets are crucial for the development of vehicle ReID in model training and evaluation. However, existing datasets in this field have limitations in many aspects, including the constrained capture conditions, limited variation of vehicle appearances, and small scale of training and test set, etc. Hence, a new, large, and challenging benchmark for vehicle ReID is urgently needed. In this paper, we propose a large vehicle ReID dataset, called VERI-Wild 2.0, containing 825,042 images. It is captured using a city-scale surveillance camera system, consisting of 274 cameras covering a very large area over 200 $km^2$ k m 2 . Specifically, the samples in our dataset present very rich appearance diversities thanks to the long time span collecting settings, unconstrained capturing viewpoints, various illumination conditions, diversified background environments, and different weather conditions. Furthermore, to facilitate more practical benchmarking, we define a challenging and large test set containing about 400K vehicle images that do not have any camera overlap with the training set. VERI-Wild 2.0 is expected to be able to facilitate the design, adaptation, development, and evaluation of different types of learning models for vehicle ReID. Besides, we also design a new method for vehicle ReID. We observe that orientation is a crucial factor for feature matching in vehicle ReID. To match vehicle pairs captured from similar orientations, the learned features are expected to capture specific detailed differential information for discriminating the visually similar yet different vehicles. In contrast, features are desired to capture the orientation invariant common information when matching samples captured from different orientations. Thus a novel disentangled feature learning network (DFNet) is proposed. It explicitly considers the orientation information for vehicle ReID, and concurrently learns the orientation specific and orientation common features that thus can be adaptively exploited via an adaptive matching scheme when dealing with matching pairs from similar or different orientations. The comprehensive experimental results show the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

This paper generalizes the Attention in Attention (AiA) mechanism, in P. Fang et al., 2019 by employing explicit mapping in reproducing kernel Hilbert spaces to generate attention values of the input feature map. The AiA mechanism models the capacity of building inter-dependencies among the local and global features by the interaction of inner and outer attention modules. Besides a vanilla AiA module, termed linear attention with AiA, two non-linear counterparts, namely, second-order polynomial attention and Gaussian attention, are also proposed to utilize the non-linear properties of the input features explicitly, via the second-order polynomial kernel and Gaussian kernel approximation. The deep convolutional neural network, equipped with the proposed AiA blocks, is referred to as Attention in Attention Network (AiA-Net). The AiA-Net learns to extract a discriminative pedestrian representation, which combines complementary person appearance and corresponding part features. Extensive ablation studies verify the effectiveness of the AiA mechanism and the use of non-linear features hidden in the feature map for attention design. Furthermore, our approach outperforms current state-of-the-art by a considerable margin across a number of benchmarks. In addition, state-of-the-art performance is also achieved in the video person retrieval task with the assistance of the proposed AiA blocks. [ABSTRACT FROM AUTHOR]

Numerous tracking approaches attempt to improve target representation through target-aware or distractor-aware. However, the unbalanced considerations of target or distractor information make it diffcult for these methods to benefit from the two aspects at the same time. In this paper, we propose a target-distractor aware model with discriminative enhancement learning loss to learn target representation, which can better distinguish the target in complex scenes. Firstly, to enlarge the gap between the target and distractor, we design a discriminative enhancement learning loss. By highlighting the hard negatives that are similar to the target and shrinking the easy negatives that are pure background, the features sensitive to the target or distractor representation can be more conveniently mined. On this basis, we further propose a target-distractor aware model. Unlike existing methods of preference target or distractor, we construct the target-specific feature space by activating the target-sensitive and the distractor-silence feature. Therefore, the appearance model can not only represent the target well but also suppress the background distractor. Finally, the target-distractor aware target representation model is integrated with a Siamese matching network for visual tracking for achieving robust and realtime visual tracking. Extensive experiments are performed on eight tracking benchmarks show that the proposed algorithm achieves favorable performance. [ABSTRACT FROM AUTHOR]

RGB-D co-salient object detection aims to segment co-occurring salient objects when given a group of relevant images and depth maps. Previous methods often adopt separate pipeline and use hand-crafted features, being hard to capture the patterns of co-occurring salient objects and leading to unsatisfactory results. Using end-to-end CNN models is a straightforward idea, but they are less effective in exploiting global cues due to the intrinsic limitation. Thus, in this paper, we alternatively propose an end-to-end transformer-based model which uses class tokens to explicitly capture implicit class knowledge to perform RGB-D co-salient object detection, denoted as CTNet. Specifically, we first design adaptive class tokens for individual images to explore intra-saliency cues and then develop common class tokens for the whole group to explore inter-saliency cues. Besides, we also leverage the complementary cues between RGB images and depth maps to promote the learning of the above two types of class tokens. In addition, to promote model evaluation, we construct a challenging and large-scale benchmark dataset, named RGBD CoSal1k, which collects 106 groups containing 1000 pairs of RGB-D images with complex scenarios and diverse appearances. Experimental results on three benchmark datasets demonstrate the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

With increasing usage of face recognition algorithms, it is well established that external artifacts and makeup accessories can be applied to different facial features such as eyes, nose, mouth, and cheek, to obfuscate one’s identity or to impersonate someone else’s identity. Recognizing faces in the presence of these artifacts comprises the problem of disguised face recognition, which is one of the most arduous covariates of face recognition. The challenge becomes exacerbated when disguised faces are captured in real-time environment, with low resolution images. To address the challenge of disguised face recognition, this paper first proposes a novel multi-objective encoder-decoder network, termed as DED-Net. DED-Net attempts to learn the class variations in the feature space generated by both disguised as well non-disguised images, using a combination of Mahalanobis and Cosine distance metrics, along with Mutual Information based supervision. The DED-Net is then extended to learn from the local and global features of both disguised and non-disguised face images for efficient face recognition, and the complete framework is termed as Disguise Resilient (D-Res) framework. The efficacy of the proposed framework has been demonstrated on two real-world benchmark datasets: Disguised Faces in the Wild (DFW) 2018 and DFW2019 competition datasets. In addition, this research also emphasizes on the importance of recognizing disguised faces in low resolution settings and proposes three experimental protocols to simulate the real-world surveillance scenario. To this effect, benchmark results have been shown on seven protocols for three low resolution settings ($32\times 32$ , $24\times 24$ , and $16\times 16$) of the two DFW benchmark datasets. The results demonstrate superior performance of the D-Res framework, in comparison with benchmark algorithms. For example, an improvement of around 3% is observed on the Overall protocol of the DFW2019 dataset, where the D-Res framework achieves 96.3%. Experiments have also been performed on benchmark face verification datasets (LFW, YTF, and IJB-B), where the D-Res framework achieves improved verification accuracy. [ABSTRACT FROM AUTHOR]

In this paper, we introduce the idea of blockwise classification to count objects. The current mainstream method for counting objects is to regress the density map or to regress the redundant count map via a deep convolutional neural network (CNN). However, these methods suffer from two critical issues: inaccurately generated regression targets and serious sample imbalances. First, the ground truth density map is generated by convolving the dot map using a Gaussian kernel. Because an inappropriate kernel can cover the background or uncover objects, this approach introduces a form of noise, and therefore results in ambiguities when training the networks. Second, inhomogeneously distributed objects often exist in images, which gives rise to a data collection bias. This leads to a long-tailed distribution of region counts, which is a typical characteristic that occurs with imbalanced samples; therefore, underestimations in high-density regions and overestimations in low-density regions are common. In this paper, we address these two issues within one framework—blockwise count level classification. The intuition behind this idea is that while it may not be possible to provide an exact count of pixels or patches, it is possible to provide a count of a region that falls within a certain interval with high confidence. Our method classifies the count levels of each block produced by nonlinearly quantizing the continuous counts, thus transforming the imbalance of sample patch counts into a class imbalance of count levels. Consequently, an information-entropy-inspired loss can be applied to alleviate this issue. Through ablative studies, we analyze the impact of imbalanced data, Gaussian kernel sizes, quantization errors, and the effectiveness of each module in our method. Without bells and whistles, our method outperforms or performs competitively with other state-of-the-art approaches on seven object-counting benchmarks, including four crowd-counting datasets from ShanghaiTech, WorldExpo’10, UCF-QNRF and UCF_CC_50, one vehicle-counting dataset (TRANCOS), one maize-tassel-counting dataset (MTC), and one challenging sonar fish-counting dataset that we constructed. The results suggest that our framework provides a strong and improved baseline for object counting. [ABSTRACT FROM AUTHOR]

Most existing work that grounds natural language phrases in images starts with the assumption that the phrase in question is relevant to the image. In this paper we address a more realistic version of the natural language grounding task where we must both identify whether the phrase is relevant to an image and localize the phrase. This can also be viewed as a generalization of object detection to an open-ended vocabulary, introducing elements of few- and zero-shot detection. We propose an approach for this task that extends Faster R-CNN to relate image regions and phrases. By carefully initializing the classification layers of our network using canonical correlation analysis (CCA), we encourage a solution that is more discerning when reasoning between similar phrases, resulting in over double the performance compared to a naive adaptation on three popular phrase grounding datasets, Flickr30K Entities, ReferIt Game, and Visual Genome, with test-time phrase vocabulary sizes of 5K, 32K, and 159K, respectively. [ABSTRACT FROM AUTHOR]

In recent years, RGBT tracking has become a hot topic in the field of visual tracking, and made great progress. In this paper, we propose a novel Trident Fusion Network (TFNet) to achieve effective fusion of different modalities for robust RGBT tracking. In specific, to deploy the complementarity of features of all convolutional layers, we propose a recursive strategy to densely aggregate these features that yield robust representations of target objects in two modalities. Moreover, we design a trident architecture to integrate the fused features and both modality-specific features for robust target representations. There are three main advantages. First, retaining the classification layer of each modality is beneficial to enhance feature learning of single modality, and compared with aggregate branches, single-modality branches pay more attention to the mining of modal specific information. Second, when some modality is noisy or invalid, the modality-specific branches would capture more discriminative features for RGBT tracking. Finally, the integration of aggregation branches and single-modality branches is beneficial to the complementary learning of different modalities. In addition, we also introduce a feature pruning module in each branch to prune the redundant features and avoid network overfitting. Experimental results on four RGBT tracking benchmark datasets suggest that our tracker achieves superior performance against the state-of-the-art RGBT tracking methods. [ABSTRACT FROM AUTHOR]

Tracking-by-detection algorithms have considerably enhanced tracking performance with the introduction of recent convolutional neural networks (CNNs). However, most trackers directly exploit standard scalar-output CNN features, which may not capture enough feature encoding information, instead of aggregated CNN features of vector-output form. In this paper, we propose an end-to-end feature aggregation capsule framework. First, based on the existing CNN network, we aggregate a certain number of similar position-aware CNN features into a capsule to model the feature similarity. The acquired vector-level feature capsules (rather than previous scalar-level pointwise features) are utilized for differentiation learning. We then propose a group attention module to better model the entity representation between different capsule groups thus optimizes total discriminative capability. Third, to reduce the prediction interference resulted by the side effect of dimension rising within capsules, we propose a penalty attention module. Such strategy could dynamically adjust values of neurons by estimating whether they are beneficial or harmful to tracking. Experimental results on five representative benchmarks (UAVDT, DTB70, UAV123, VOT2016 and VOT2018) demonstrate the excellent tracking performance of our dual attention based capsule tracker (DACapT). Specially, it exceeds the previous top tracker by 4.6%/1.9% in precision/success evaluations on UAVDT. [ABSTRACT FROM AUTHOR]

Fault diagnosis of bearings using classification techniques plays an important role in industrial applications, and, hence, has received increasing attention. Recently, significant efforts have been made to develop various methods for bearing fault classification and the application of Case Western Reserve University (CWRU) data for validation has become a standard reference to test the fault classification algorithms. However, a systematic research for evaluating bearing fault classification performance using the CWRU data is still lacking. This paper aims to provide a comprehensive benchmark analysis of the CWRU data using various entropy and classification methods. The main contribution of this paper is applying entropy-based fault classification methods to establish a benchmark analysis of entire CWRU datasets, aiming to provide a proper assessment of any new classification methods. Recommendations are provided for the selection of the CWRU data to aid in testing new fault classification algorithms, which will enable the researches to develop and evaluate various diagnostic algorithms. In the end, the comparison results and discussion are reported as a useful baseline for future research. [ABSTRACT FROM AUTHOR]

This paper proposes a generic method to learn interpretable convolutional filters in a deep convolutional neural network (CNN) for object classification, where each interpretable filter encodes features of a specific object part. Our method does not require additional annotations of object parts or textures for supervision. Instead, we use the same training data as traditional CNNs. Our method automatically assigns each interpretable filter in a high conv-layer with an object part of a certain category during the learning process. Such explicit knowledge representations in conv-layers of the CNN help people clarify the logic encoded in the CNN, i.e., answering what patterns the CNN extracts from an input image and uses for prediction. We have tested our method using different benchmark CNNs with various architectures to demonstrate the broad applicability of our method. Experiments have shown that our interpretable filters are much more semantically meaningful than traditional filters. [ABSTRACT FROM AUTHOR]

RGB-D salient object detection (SOD) aims to segment the most attractive objects in a pair of cross-modal RGB and depth images. Currently, most existing RGB-D SOD methods focus on the foreground region when utilizing the depth images. However, the background also provides important information in traditional SOD methods for promising performance. To better explore salient information in both foreground and background regions, this paper proposes a Bilateral Attention Network (BiANet) for the RGB-D SOD task. Specifically, we introduce a Bilateral Attention Module (BAM) with a complementary attention mechanism: foreground-first (FF) attention and background-first (BF) attention. The FF attention focuses on the foreground region with a gradual refinement style, while the BF one recovers potentially useful salient information in the background region. Benefited from the proposed BAM module, our BiANet can capture more meaningful foreground and background cues, and shift more attention to refining the uncertain details between foreground and background regions. Additionally, we extend our BAM by leveraging the multi-scale techniques for better SOD performance. Extensive experiments on six benchmark datasets demonstrate that our BiANet outperforms other state-of-the-art RGB-D SOD methods in terms of objective metrics and subjective visual comparison. Our BiANet can run up to 80 fps on $224\times 224$ RGB-D images, with an NVIDIA GeForce RTX 2080Ti GPU. Comprehensive ablation studies also validate our contributions. [ABSTRACT FROM AUTHOR]

Detecting incidentally captured text in the wild remains an open problem due to challenging factors including unconstrained scenarios and large scale variation. In this paper, we establish a large-scale scene text detection dataset (LS-Text), containing 36, 000 images and 270, 783 text instances with various scales and complex scenarios, to promote the research of text detection. We propose a Scale-residual Learning Network (SLN) to deal with the scale variation problem in a progressive optimization manner. Specifically, we integrate both learnable feature concatenation and feature up-sampling operator. It can effectively eliminate the residuals between the outputs of SLN and ground-truth text instances by processing both the Feature Fusion Residuals (FFR) and the Scale Transformation Residuals (STR), simultaneously. By stacking multi-scale feature maps in a deep-to-shallow manner, SLN continuously optimizes feature representation by accumulating strong semantic information and rich texture details in a scale-residual learning way. Extensive experimental results on five challenging datasets demonstrate the state-of-the-art performance of the proposed SLN model, and the challenging aspects related to real-world scenarios of the proposed LS-Text dataset. Both the source code of SLN and the LS-Text dataset are available at https://github.com/SLN-Text-Detection. [ABSTRACT FROM AUTHOR]

In this paper, we report the results of the first international contest on mitosis detection in phase-contrast microscopy image sequences (https://www.iti-tju.org/mitosisdetection), which was held at the workshop of computer vision for microscopy image analysis (CVMI) in CVPR 2019. This contest aims to promote research on spatiotemporal mitosis detection under microscopy images. In this contest, we released a large-scale time-lapse phase-contrast microscopy image dataset (C2C12-16) for the mitosis detection task. Compared with the previous popular datasets (e.g., C2C12, C3H10), C2C12-16 contains more annotated mitotic events and more diverse cell culture environments. A total of ten different mitosis detection methods were submitted in the contest and evaluated on the test sets of four different cell culture environments in C2C12-16. In this benchmark, we describe all methods and conduct a thorough analysis based on their performances and discuss a feasible direction for mitosis detection. To the best of our knowledge, this is the first benchmark for the mitosis detection problem using a time-lapse phase-contrast microscopy spatiotemporal image sequence model. [ABSTRACT FROM AUTHOR]

Object detection has recently experienced substantial progress. Yet, the widely adopted horizontal bounding box representation is not appropriate for ubiquitous oriented objects such as objects in aerial images and scene texts. In this paper, we propose a simple yet effective framework to detect multi-oriented objects. Instead of directly regressing the four vertices, we glide the vertex of the horizontal bounding box on each corresponding side to accurately describe a multi-oriented object. Specifically, We regress four length ratios characterizing the relative gliding offset on each corresponding side. This may facilitate the offset learning and avoid the confusion issue of sequential label points for oriented objects. To further remedy the confusion issue for nearly horizontal objects, we also introduce an obliquity factor based on area ratio between the object and its horizontal bounding box, guiding the selection of horizontal or oriented detection for each object. We add these five extra target variables to the regression head of faster R-CNN, which requires ignorable extra computation time. Extensive experimental results demonstrate that without bells and whistles, the proposed method achieves superior performances on multiple multi-oriented object detection benchmarks including object detection in aerial images, scene text detection, pedestrian detection in fisheye images. [ABSTRACT FROM AUTHOR]

The performances of person re-identification tasks can be seriously degraded because of variations caused by view changes. In recent years, there are many methods focusing on how to solve cross view challenges which can be roughly divided into two categories: 1) learning view-invariant features without the help of view information. 2) combining view-wise features with the guide of view information. However, these methods are neither perfect enough. Methods of the first category are not roust enough for different kinds of view-invariants while methods of the other category can not generalize well in real-world applications. In this paper, we aim to learn view-invariant features with the help of view information. We proposed an end-to-end trainable framework, called View Confusion Feature Learning (VCFL), to learn view-invariant features by getting rid of view specific information. To the best of our knowledge, VCFL is originally proposed to learn view-invariant identity-wise features, and it is a kind of combination of view-generic and view-specific methods. The whole view confusion learning mechanism consists of three parts: 1) adversarial learning between feature extractor and the view classifier; 2) drawing the features with the same ID close to centers; 3) the guidance of SIFT, for seamlessly integration of hand-crafted features and deep features. In order to make the whole confusion mechanism work better, we further propose a VCFL+ model, which improves the fusion process in the feature map level through the thoughts of attention mechanism. Experiments on three benchmark datasets including Market1501, CUHK03, and DukeMTMC prove the superiority of our method over state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Saliency detection is a fundamental and challenging task in computer vision, which aims at distinguishing the most conspicuous objects or regions in an image. Existing deep-learning methods mainly rely on the entire image to learn the global context information for saliency detection, which loses the spatial relation and results in ambiguity in predicting saliency maps. In this paper, we propose a novel deep sub-region network (DSR-Net) equipped with a sequence of sub-region dilated blocks (SRDB) by aggregating multi-scale salient context information of multiple sub-regions, such that the global context information from the whole image and local contexts from sub-regions are fused together, making the saliency prediction more accurate. Our SRDB separates the input feature map at different layers of a convolutional neural network (CNN) into different sub-regions and then designs a parallel ASPP module to refine feature maps at each sub-region. Experiments on the five widely-used saliency benchmark datasets demonstrate that our network outperforms recent state-of-the-art saliency detectors quantitatively and qualitatively on all the benchmarks. [ABSTRACT FROM AUTHOR]

Object proposals are used in two-stage detectors, such as R-CNN, to generate detection results, including category predictions and refined bounding-boxes. As a result, classification scores are assigned to refined bounding-boxes rather than object proposals. However, this procedure ignores the discrepancy of data distribution between object proposals and refined bounding-boxes. We consider this discrepancy could limit the detection accuracy. Specifically, the foreground/background imbalance on object proposals and inaccurate information from low-IoU proposals could hinder the category prediction. In this paper, we propose a detector called the Multi-Path Detection Calibration Network (PDC-Net) to address this problem. The key idea behind PDC-Net is calibrating detection results from R-CNN by considering the statistical discrepancy between object proposals and refined bounding-boxes. PDC-Net is built on Faster R-CNN. The core component in PDC-Net is the multi-path detection head, in which the base detector (from Faster R-CNN) generates detection results from object proposals and multiple calibration detectors fix incorrect outputs from the base detector using refined bounding-boxes. Experiments reveal that PDC-Net can boost detection results. Our method could reach 83.1% and 43.3% mAP respectively on PASCAL VOC and MSCOCO benchmarks, which is comparable to several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Salient object detection is a problem that has been considered in detail and many solutions have been proposed. In this paper, we argue that work to date has addressed a problem that is relatively ill-posed. Specifically, there is not universal agreement about what constitutes a salient object when multiple observers are queried. This implies that some objects are more likely to be judged salient than others, and implies a relative rank exists on salient objects. Initially, we present a novel deep learning solution based on a hierarchical representation of relative saliency and stage-wise refinement. Further to this, we present data, analysis and baseline benchmark results towards addressing the problem of salient object ranking. Methods for deriving suitable ranked salient object instances are presented, along with metrics suitable to measuring algorithm performance. In addition, we show how a derived dataset can be successively refined to provide cleaned results that correlate well with pristine ground truth in its characteristics and value for training and testing models. Finally, we provide a comparison among prevailing algorithms that address salient object ranking or detection to establish initial baselines providing a basis for comparison with future efforts addressing this problem. The source code and data are publicly available via our project page: ryersonvisionlab.github.io/cocosalrank. [ABSTRACT FROM AUTHOR]

The key to realize advanced document summarization is semantic representation of documents. This paper investigates the role of Semantic Link Network in representing and understanding documents for multi-document summarization. It proposes a novel abstractive multi-document summarization framework by first transforming documents into a Semantic Link Network of concepts and events and then transforming the Semantic Link Network into the summary of the documents based on the selection of important concepts and events while keeping semantics coherence. Experiments on benchmark datasets show that the proposed summarization approach significantly outperforms relevant state-of-the-art baselines and the Semantic Link Network plays an important role in representing and understanding documents. [ABSTRACT FROM AUTHOR]

Shadow detection and shadow removal are fundamental and challenging tasks, requiring an understanding of the global image semantics. This paper presents a novel deep neural network design for shadow detection and removal by analyzing the spatial image context in a direction-aware manner. To achieve this, we first formulate the direction-aware attention mechanism in a spatial recurrent neural network (RNN) by introducing attention weights when aggregating spatial context features in the RNN. By learning these weights through training, we can recover direction-aware spatial context (DSC) for detecting and removing shadows. This design is developed into the DSC module and embedded in a convolutional neural network (CNN) to learn the DSC features at different levels. Moreover, we design a weighted cross entropy loss to make effective the training for shadow detection and further adopt the network for shadow removal by using a euclidean loss function and formulating a color transfer function to address the color and luminosity inconsistencies in the training pairs. We employed two shadow detection benchmark datasets and two shadow removal benchmark datasets, and performed various experiments to evaluate our method. Experimental results show that our method performs favorably against the state-of-the-art methods for both shadow detection and shadow removal. [ABSTRACT FROM AUTHOR]

In this paper, a novel benchmark is introduced for evaluating local image descriptors. We demonstrate limitations of the commonly used datasets and evaluation protocols, that lead to ambiguities and contradictory results in the literature. Furthermore, these benchmarks are nearly saturated due to the recent improvements in local descriptors obtained by learning from large annotated datasets. To address these issues, we introduce a new large dataset suitable for training and testing modern descriptors, together with strictly defined evaluation protocols in several tasks such as matching, retrieval and verification. This allows for more realistic, thus more reliable comparisons in different application scenarios. We evaluate the performance of several state-of-the-art descriptors and analyse their properties. We show that a simple normalisation of traditional hand-crafted descriptors is able to boost their performance to the level of deep learning based descriptors once realistic benchmarks are considered. Additionally we specify a protocol for learning and evaluating using cross validation. We show that when training state-of-the-art descriptors on this dataset, the traditional verification task is almost entirely saturated. [ABSTRACT FROM AUTHOR]

Direct regression methods have demonstrated their success on various multi-oriented benchmarks for scene text detection due to the high recall rate for small targets and the direct regression for text boxes. However, too many false positive candidates and inaccurate position regression still limit the performance of these methods. In this paper, we propose an end-to-end method by introducing position-sensitive segmentation into the direct regression method to overcome these shortcomings. We generate the ground truth of position-sensitive segmentation maps based on the information of text boxes so that the position-sensitive segmentation module can be trained synchronously with the direct regression module. Besides, more information about the relative position of text is provided for the network through the training of position-sensitive segmentation maps, which improves the expressiveness of the network. We also introduce spatial pyramid of position-sensitive segmentation into the proposed method considering the huge differences in sizes and aspect ratios of scene texts and we propose position-sensitive COI(Corner area of Interest) pooling into the proposed method to speed up the inference. Experiments on datasets ICDAR2015, MLT-17 and COCO-Text demonstrate that the proposed method has a comparable performance with state-of-the-art methods while it is more efficient. We also provide abundant ablation experiments to demonstrate the effectiveness of these improvements in our proposed method. [ABSTRACT FROM AUTHOR]

In this paper, we study the challenging problem of categorizing videos according to high-level semantics such as the existence of a particular human action or a complex event. Although extensive efforts have been devoted in recent years, most existing works combined multiple video features using simple fusion strategies and neglected the utilization of inter-class semantic relationships. This paper proposes a novel unified framework that jointly exploits the feature relationships and the class relationships for improved categorization performance. Specifically, these two types of relationships are estimated and utilized by imposing regularizations in the learning process of a deep neural network (DNN). Through arming the DNN with better capability of harnessing both the feature and the class relationships, the proposed regularized DNN (rDNN) is more suitable for modeling video semantics. We show that rDNN produces better performance over several state-of-the-art approaches. Competitive results are reported on the well-known Hollywood2 and Columbia Consumer Video benchmarks. In addition, to stimulate future research on large scale video categorization, we collect and release a new benchmark dataset, called FCVID, which contains 91,223 Internet videos and 239 manually annotated categories. [ABSTRACT FROM PUBLISHER]

This paper presents a novel deep learning model to aggregate the attentional dilated features for salient object detection by exploring the complementary information between the global and local context in a convolutional neural network. There are two technical contributions to our network design. First, we develop an attentional dense atrous (dilated) spatial pyramid pooling (AD-ASPP) module to selectively use the local saliency cues captured by dilated convolutions with a small rate and the global saliency cues captured by dilated convolutions with a large rate. Second, taking the feature pyramid network as the backbone, we develop an aggregation network to integrate the refined features by formulating two consecutive chains of residual learning based modules: one chain from deep to shallow layers while another chain from shallow to deep layers. We evaluate our network on seven widely-used saliency detection benchmarks by comparing it against 21 state-of-the-art methods. Experimental results show that our network outperforms others on all the seven benchmark datasets. [ABSTRACT FROM AUTHOR]

Crowd counting is challenging due to unconstrained imaging factors, e.g., background clutters, non-uniform distribution of people, large scale and perspective variations. Dealing with these problems using deep neural networks requires rich prior knowledge and multi-scale contextual representations. In this paper, we propose a Cross-stage Refinement Network (CRNet) that can refine predicted density maps progressively based on hierarchical multi-level density priors. In particular, CRNet is composed of several fully convolutional networks. They are stacked together recursively with the previous output as the next input, and each of them serves to utilize previous density output to gradually correct prediction errors of crowd areas and refine the predicted density maps at different stages. Cross-stage multi-level density priors are further exploited in our recurrent framework by the cross-stage skip layers based on ConvLSTM. To cope with different challenges of unconstrained crowd scenes, we explore different crowd-specific data augmentation methods to mimic real-world scenarios and enrich crowd feature representations from different aspects. Extensive experiments show the proposed method achieves superior performances against state-of-the-art methods on four widely-used challenging benchmarks in terms of counting accuracy and density map quality. Code and models are available at this https://github.com/lytgftyf/Crowd-Counting-via-Cross-stage-Refinement-Networks. [ABSTRACT FROM AUTHOR]

Deep trackers have proven success in visual tracking. Typically, these trackers employ optimally pre-trained deep networks to represent all diverse objects with multi-channel features from some fixed layers. The deep networks employed are usually trained to extract rich knowledge from massive data used in object classification and so they are capable to represent generic objects very well. However, these networks are too complex to represent a specific moving object, leading to poor generalization as well as high computational and memory costs. This paper presents a novel and general framework termed channel distillation to facilitate deep trackers. To validate the effectiveness of channel distillation, we take discriminative correlation filter (DCF) and ECO for example. We demonstrate that an integrated formulation can turn feature compression, response map generation, and model update into a unified energy minimization problem to adaptively select informative feature channels that improve the efficacy of tracking moving objects on the fly. Channel distillation can accurately extract good channels, alleviating the influence of noisy channels and generally reducing the number of channels, as well as adaptively generalizing to different channels and networks. The resulting deep tracker is accurate, fast, and has low memory requirements. Extensive experimental evaluations on popular benchmarks clearly demonstrate the effectiveness and generalizability of our framework. [ABSTRACT FROM AUTHOR]

In this paper, we explore how to leverage readily available unlabeled data to improve semi-supervised human detection performance. For this purpose, we specifically modify the region proposal network (RPN) for learning on a partially labeled dataset. Based on commonly observed false positive types, a verification module is developed to assess foreground human objects in the candidate regions to provide an important cue for filtering the RPN’s proposals. The remaining proposals with high confidence scores are then used as pseudo annotations for re-training our detection model. To reduce the risk of error propagation in the training process, we adopt a self-paced training strategy to progressively include more pseudo annotations generated by the previous model over multiple training rounds. The resulting detector re-trained on the augmented data can be expected to have better detection performance. The effectiveness of the main components of this framework is verified through extensive experiments, and the proposed approach achieves state-of-the-art detection results on multiple scene-specific human detection benchmarks in the semi-supervised setting. [ABSTRACT FROM AUTHOR]

Person re-identification (re-ID) has been a popular topic in computer vision and pattern recognition communities for a decade. Several important milestones such as metric-based and deeply-learned re-ID in recent years have promoted this topic. However, most existing re-ID works are designed for closed-world scenarios rather than realistic open-world settings, which limits the practical application of the re-ID technique. On one hand, the performance of the latest re-ID methods has surpassed the human-level performance on several commonly used benchmarks (e.g., Market1501 and CUHK03), which are collected from closed-world scenarios. On the other hand, open-world tasks that are less developed and more challenging have received increasing attention in the re-ID community. Therefore, this paper starts the first attempt to analyze the trends of open-world re-ID and summarizes them from both narrow and generalized perspectives. In the narrow perspective, open-world re-ID is regarded as person verification (i.e., open-set re-ID) instead of person identification, that is, the query person may not occur in the gallery set. In the generalized perspective, application-driven methods that are designed for specific applications are defined as generalized open-world re-ID. Their settings are usually close to realistic application requirements. Specifically, this survey mainly includes the following four points for open-world re-ID: 1) analyzing the discrepancies between closed- and open-world scenarios; 2) describing the developments of existing open-set re-ID works and their limitations; 3) introducing specific application-driven works from three aspects, namely, raw data, practical procedure, and efficiency; and 4) summarizing the state-of-the-art methods and future directions for open-world re-ID. This survey on open-world re-ID provides a guidance for improving the usability of re-ID technique in practical applications. [ABSTRACT FROM AUTHOR]

Discriminative correlation filter (DCF)-based trackers have recently exhibited high efficiency and impressive robustness to challenging factors, such as illumination change and partial occlusion. However, in cases with fast motion and full occlusion, these trackers drift off soon and can hardly re-detect the target from the restricted search region due to the boundary effect. On the contrary, recent work using a fully convolutional Siamese network (Siamfc) locates the exemplar image within a large search image but suffers from coarse location and distractors. In this paper, we propose a real-time complementary tracker (RCT) by integrating DCF and Siamfc into a two-stage tracking framework where DCF and Siamfc share mutual advantages and complement each other. In the first stage of this framework, RCT locates the target coarsely but robustly with Siamfc. In the second stage, the derived coarse location is refined by DCF for higher accuracy. For efficiency reasons, Siamfc in the first stage is activated occasionally based on the tracking status inferred from the correlation response map of DCF in the second stage. Comprehensive experiments are performed on three popular benchmark datasets: OTB2013, OTB2015, and VOT2016. On OTB2013, RCT runs with over 40 f/s and achieves an absolute gain of 4.8% and 5.2% in mean overlap precision compared with two base trackers (Staple and Siamfc). On VOT2016, RCT makes a good balance between performance and efficiency, ranking fifth in EAO and first in EFO compared with the top five trackers. [ABSTRACT FROM AUTHOR]

Segmentation of cardiac ventricle from magnetic resonance images is significant for cardiac disease diagnosis, progression assessment, and monitoring cardiac conditions. Manual segmentation is so time consuming, tedious, and subjective that automated segmentation methods are highly desired in practice. However, conventional segmentation methods performed poorly in cardiac ventricle, especially in the right ventricle. Compared with the left ventricle, whose shape is a simple thick-walled circle, the structure of the right ventricle is more complex due to ambiguous boundary, irregular cavity, and variable crescent shape. Hence, effective feature extractors and segmentation models are preferred. In this paper, we propose a dilated-inception net (DIN) to extract and aggregate multi-scale features for right ventricle segmentation. The DIN outperforms many state-of-the-art models on the benchmark database of right ventricle segmentation challenge. In addition, the experimental results indicate that the proposed model has potential to reach expert-level performance in right ventricular epicardium segmentation. More importantly, DIN behaves similarly to clinical expert with high correlation coefficients in four clinical cardiac indices. Therefore, the proposed DIN is promising for automated cardiac right ventricle segmentation in clinical applications. [ABSTRACT FROM AUTHOR]

Big data has had a great share in the success of deep learning in computer vision. Recent works suggest that there is significant further potential to increase object detection performance by utilizing even bigger datasets. In this paper, we introduce the EuroCity Persons dataset, which provides a large number of highly diverse, accurate and detailed annotations of pedestrians, cyclists and other riders in urban traffic scenes. The images for this dataset were collected on-board a moving vehicle in 31 cities of 12 European countries. With over 238,200 person instances manually labeled in over 47,300 images, EuroCity Persons is nearly one order of magnitude larger than datasets used previously for person detection in traffic scenes. The dataset furthermore contains a large number of person orientation annotations (over 211,200). We optimize four state-of-the-art deep learning approaches (Faster R-CNN, R-FCN, SSD and YOLOv3) to serve as baselines for the new object detection benchmark. In experiments with previous datasets we analyze the generalization capabilities of these detectors when trained with the new dataset. We furthermore study the effect of the training set size, the dataset diversity (day- versus night-time, geographical region), the dataset detail (i.e., availability of object orientation information) and the annotation quality on the detector performance. Finally, we analyze error sources and discuss the road ahead. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel visual-inertial odometry (VIO) approach that adopts structural regularity in man-made environments. Instead of using Manhattan world assumption, we use Atlanta world model to describe such regularity. An Atlanta world is a world that contains multiple local Manhattan worlds with different heading directions. Each local Manhattan world is detected on the fly, and their headings are gradually refined by the state estimator when new observations are received. With full exploration of structural lines that aligned with each local Manhattan worlds, our VIO method becomes more accurate and robust, as well as more flexible to different kinds of complex man-made environments. Through benchmark tests and real-world tests, the results show that the proposed approach outperforms existing visual-inertial systems in large-scale man-made environments. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation aims to leverage the labeled source data to learn with the unlabeled target data. Previous trandusctive methods tackle it by iteratively seeking a low-dimensional projection to extract the invariant features and obtaining the pseudo target labels via building a classifier on source data. However, they merely concentrate on minimizing the cross-domain distribution divergence, while ignoring the intra-domain structure especially for the target domain. Even after projection, possible risk factors like imbalanced data distribution may still hinder the performance of target label inference. In this paper, we propose a simple yet effective domain-invariant projection ensemble approach to tackle these two issues together. Specifically, we seek the optimal projection via a novel relaxed domain-irrelevant clustering-promoting term that jointly bridges the cross-domain semantic gap and increases the intra-class compactness in both domains. To further enhance the target label inference, we first develop a ‘sampling-and-fusion’ framework, under which multiple projections are independently learned based on various randomized coupled domain subsets. Subsequently, aggregating models such as majority voting are utilized to leverage multiple projections and classify unlabeled target data. Extensive experimental results on six visual benchmarks including object, face, and digit images, demonstrate that the proposed methods gain remarkable margins over state-of-the-art unsupervised domain adaptation methods. [ABSTRACT FROM AUTHOR]

Retrieving specific persons with various types of queries, e.g., a set of attributes or a portrait photo has great application potential in large-scale intelligent surveillance systems. In this paper, we propose a richly annotated pedestrian (RAP) dataset which serves as a unified benchmark for both attribute-based and image-based person retrieval in real surveillance scenarios. Typically, previous datasets have three improvable aspects, including limited data scale and annotation types, heterogeneous data source, and controlled scenarios. Differently, RAP is a large-scale dataset which contains 84928 images with 72 types of attributes and additional tags of viewpoint, occlusion, body parts, and 2589 person identities. It is collected in the real uncontrolled scene and has complex visual variations in pedestrian samples due to the change of viewpoints, pedestrian postures, and cloth appearance. Towards a high-quality person retrieval benchmark, an amount of state-of-the-art algorithms on pedestrian attribute recognition and person re-identification (ReID), are performed for quantitative analysis with three evaluation tasks, i.e., attribute recognition, attribute-based and image-based person retrieval, where a new instance-based metric is proposed to measure the dependency of the prediction of multiple attributes. Finally, some interesting problems, e.g., the joint feature learning of attribute recognition and ReID, and the problem of cross-day person ReID, are explored to show the challenges and future directions in person retrieval. [ABSTRACT FROM AUTHOR]

Recent advances in saliency detection have utilized deep learning to obtain high-level features to detect salient regions in scenes. These advances have yielded results superior to those reported in past work, which involved the use of hand-crafted low-level features for saliency detection. In this paper, we propose ELD-Net, a unified deep learning framework for accurate and efficient saliency detection. We show that hand-crafted features can provide complementary information to enhance saliency detection that uses only high-level features. Our method uses both low-level and high-level features for saliency detection. High-level features are extracted using GoogLeNet, and low-level features evaluate the relative importance of a local region using its differences from other regions in an image. The two feature maps are independently encoded by the convolutional and the ReLU layers. The encoded low-level and high-level features are then combined by concatenation and convolution. Finally, a linear fully connected layer is used to evaluate the saliency of a queried region. A full resolution saliency map is obtained by querying the saliency of each local region of an image. Since the high-level features are encoded at low resolution, and the encoded high-level features can be reused for every query region, our ELD-Net is very fast. Our experiments show that our method outperforms state-of-the-art deep learning-based saliency detection methods. [ABSTRACT FROM AUTHOR]

This paper proposes a method for line segment detection in digital images. We propose a novel linelet-based representation to model intrinsic properties of line segments in rasterized image space. Based on this, line segment detection, validation, and aggregation frameworks are constructed. For a numerical evaluation on real images, we propose a new benchmark dataset of real images with annotated lines called YorkUrban-LineSegment. The results show that the proposed method outperforms state-of-the-art methods numerically and visually. To our best knowledge, this is the first report of numerical evaluation of line segment detection on real images. [ABSTRACT FROM PUBLISHER]

This paper introduces matrix product state (MPS) decomposition as a new and systematic method to compress multidimensional data represented by higher order tensors. It solves two major bottlenecks in tensor compression: computation and compression quality. Regardless of tensor order, MPS compresses tensors to matrices of moderate dimension, which can be used for classification. Mainly based on a successive sequence of singular value decompositions, MPS is quite simple to implement and arrives at the global optimal matrix, bypassing local alternating optimization, which is not only computationally expensive but cannot yield the global solution. Benchmark results show that MPS can achieve better classification performance with favorable computation cost compared to other tensor compression methods. [ABSTRACT FROM PUBLISHER]

In this paper, we propose a rotation-invariant local binary descriptor (RI-LBD) learning method for visual recognition. Compared with hand-crafted local binary descriptors, such as local binary pattern and its variants, which require strong prior knowledge, local binary feature learning methods are more efficient and data-adaptive. Unlike existing learning-based local binary descriptors, such as compact binary face descriptor and simultaneous local binary feature learning and encoding, which are susceptible to rotations, our RI-LBD first categorizes each local patch into a rotational binary pattern (RBP), and then jointly learns the orientation for each pattern and the projection matrix to obtain RI-LBDs. As all the rotation variants of a patch belong to the same RBP, they are rotated into the same orientation and projected into the same binary descriptor. Then, we construct a codebook by a clustering method on the learned binary codes, and obtain a histogram feature for each image as the final representation. In order to exploit higher order statistical information, we extend our RI-LBD to the triple rotation-invariant co-occurrence local binary descriptor (TRICo-LBD) learning method, which learns a triple co-occurrence binary code for each local patch. Extensive experimental results on four different visual recognition tasks, including image patch matching, texture classification, face recognition, and scene classification, show that our RI-LBD and TRICo-LBD outperform most existing local descriptors. [ABSTRACT FROM AUTHOR]

Recent progresses in multimedia event detection have enabled us to find videos about a predefined event from a large-scale video collection. Research towards more intrinsic unsupervised video understanding is an interesting but understudied field. Specifically, given a collection of videos sharing a common event of interest, the goal is to discover the salient fragments, i.e., the curt video fragments that can concisely portray the underlying event of interest, from each video. To explore this novel direction, this paper proposes an unsupervised event saliency revealing framework. It first extracts features from multiple modalities to represent each shot in the given video collection. Then, these shots are clustered to build the cluster-level event saliency revealing framework, which explores useful information cues (i.e., the intra-cluster prior, inter-cluster discriminability, and inter-cluster smoothness) by a concise optimization model. Compared with the existing methods, our approach could highlight the intrinsic stimulus of the unseen event within a video in an unsupervised fashion. Thus, it could potentially benefit to a wide range of multimedia tasks like video browsing, understanding, and search. To quantitatively verify the proposed method, we systematically compare the method to a number of baseline methods on the TRECVID benchmarks. Experimental results have demonstrated its effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

Automatic violence detection from video is a hot topic for many video surveillance applications. However, there has been little success in developing an algorithm that can detect violence in surveillance videos with high performance. In this paper, following our recently proposed idea of motion Weber local descriptor (WLD), we make two major improvements and propose a more effective and efficient algorithm for detecting violence from motion images. First, we propose an improved WLD (IWLD) to better depict low-level image appearance information, and then extend the spatial descriptor IWLD by adding a temporal component to capture local motion information and hence form the motion IWLD. Second, we propose a modified sparse-representation-based classification model to both control the reconstruction error of coding coefficients and minimize the classification error. Based on the proposed sparse model, a class-specific dictionary containing dictionary atoms corresponding to the class labels is learned using class labels of training samples. With this learned dictionary, not only the representation residual but also the representation coefficients become discriminative. A classification scheme integrating the modified sparse model is developed to exploit such discriminative information. The experimental results on three benchmark data sets have demonstrated the superior performance of the proposed approach over the state of the arts. [ABSTRACT FROM PUBLISHER]

Multimedia data has become a major data type in the Big Data era. The explosive volume of such data and the increasing real-time requirement to retrieve useful information from it have put significant pressure in processing such data in a timely fashion. However, while prior efforts have done in-depth analysis on architectural characteristics of traditional multimedia processing and text-based retrieval algorithms, there has been no systematic study towards the emerging multimedia retrieval applications. This may impede the architecture design and system evaluation of these applications. In this paper, we make the first attempt to construct a multimedia retrieval benchmark suite (MMRBench for short) that can be used to evaluate architectures and system designs for multimedia retrieval applications. MMRBench covers modern multimedia retrieval algorithms with different versions (sequential, parallel and distributed). MMRBench also provides a series of flexible interfaces as well as certain automation tools. With such a flexible design, the algorithms in MMRBench can be used both in individual kernel-level evaluation and in integration to form a complete multimedia data retrieval infrastructure for full system evaluation. Furthermore, we use performance counters to analyze a set of architecture characteristics of multimedia retrieval algorithms in MMRBench, including the characteristics of core level, chip level and inter-chip level. The study shows that micro-architecture design in current processor is inefficient (both in performance and power) for these multimedia retrieval workloads, especially in core resources and memory systems. We then derive some insights into the architecture design and system evaluation for such multimedia retrieval algorithms. [ABSTRACT FROM AUTHOR]

Algorithms for the estimation of epipolar geometry from a pair of images have been very successful in dealing with challenging wide baseline images. In this paper the problem of scenes with repeated structures is addressed, dealing with the common case where the overlap between the images consists mainly of facades of a building. These facades may contain many repeated structures that can not be matched locally, causing state-of-the-art algorithms to fail. Assuming that the repeated structures lie on a planar surface in an ordered fashion the goal is to match them. Our algorithm first rectifies the images such that the facade is fronto-parallel. It then clusters similar features in each of the two images and matches the clusters. From them a set of hypothesized homographies of the facade is generated, using local groups of features. For each homography the epipole is recovered, yielding a fundamental matrix. For the best solution, it then decides whether the fundamental matrix has been recovered reliably and, if not, returns only the homography. The algorithm has been tested on a large number of challenging image pairs of buildings from the benchmark ZuBuD database, outperforming several state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]