Search

Unsupervised domain adaptive person re-identification aims to solve the problem of poor performance caused by transferring an unlabeled target domain from the labeled source domain in the re-identification task. The clustering pseudo-labels method in unsupervised learning is widely used in unsupervised adaptive person re-identification tasks, and it maintains state-of-the-art performance. However, pseudo-labels obtained through clustering often have much noise, and the use of a single network model structure and a single clustering algorithm can easily cause model learning to stagnate, making the model not generalizable. To solve this problem, this paper proposes an asymmetric mutual mean-teaching method for unsupervised adaptive person re-identification. In terms of feature extraction, two asymmetric network models with different structures are used for mutual mean-teaching on the target domain, making the features extracted by the network more robust. In terms of feature clustering, two clustering methods are used for mutual teaching to dynamically update the centroid of clusters to improve the confidence of clustering pseudo-labels. Finally, the triplet loss is improved based on the updated cluster centroid to improve the clustering effect. The proposed method is used to perform a large number of verification experiments on three public datasets. The experimental results show that the proposed method has better accuracy than other unsupervised person re-identification based on clustering pseudo-labels.

Neural networks have enabled state-of-the-art approaches to achieve incredible results on computer vision tasks such as object detection. However, previous works have tried to improve the performance in various object detection necks but have failed to extract features efficiently. To solve the insufficient features of objects, this work introduces some of the most advanced and representative network models based on the Faster R-CNN architecture, such as Libra R-CNN, Grid R-CNN, guided anchoring, and GRoIE. We observed the performance of Neighbour Feature Pyramid Network (NFPN) fusion, ResNet Region of Interest Feature Extraction (ResRoIE) and the Recursive Feature Pyramid (RFP) architecture at different scales of precision when these components were used in place of the corresponding original members in various networks obtained on the MS COCO dataset. Compared to the experimental results after replacing the neck and RoIE parts of these models with our Reinforced Neighbour Feature Fusion (RNFF) model, the average precision (AP) is increased by 3.2 percentage points concerning the performance of the baseline network.

With the development of science and technology, artificial intelligence has been widely used in the transportation field, and research on the symmetry of artificial intelligence has become increasingly more in-depth. Traffic sign detection based on deep learning has the problems of different target shapes and high variability in the number of targets between different labels. To solve these problems from a lack of symmetry, the idea of applying the concept of balanced data and the deformable positioning region to a target recognition network is proposed. The research is based on the improvement of the Libra R-CNN. Aiming at the problem that the difficult-to-distinguish target in target detection has a high impact on detection, the idea of generating increasingly more diverse indistinguishable samples during training is proposed to improve the detection accuracy, which is verified by experiments. The experiment is carried out on the MS COCO 2017 and traffic sign datasets. The improved Libra R-CNN is 3 percentage points better than the unimproved Libra R-CNN's mean Average Precision (mAP). A large number of comparative experimental results show that the improved network is effective.

Recent deep learning-based methods for saliency detection have proved the effectiveness of integrating features with different scales. They usually design various complex architectures of network, e.g., multiple networks, to explore the multi-scale information of images, which is expensive in computation and memory. Feature maps produced with different subsampling convolutional layers have different spatial resolutions; therefore, they can be used as the multi-scale features to reduce the costs. In this paper, by exploiting the in-network feature hierarchy of convolutional networks, we propose a novel multi-scale network for saliency detection (MSNSD) consisting of three modules, i.e., bottom-up feature extraction, top-down feature connection and multi-scale saliency prediction. Moreover, to further boost the performance of MSNSD, an input image-aware saliency aggregation method is proposed based on the ridge regression, which combines MSNSD with some well-performed handcrafted shallow models. Extensive experiments on several benchmarks show that the proposed MSNSD outperforms the state-of-the-art saliency methods with less computational and memory complexity. Meanwhile, our aggregation method for saliency detection is effective and efficient to combine deep and shallow models and make them complementary to each other.

In order to better solve the problem about lack of real time in road region segmentation algorithm, this paper designs a deep neural network of road region segmentation Spp-U-net with U-net structure as the core. In addition, the use of spatial pyramid pooling (Spp) to replace the conventional pooling makes up for the lack of accuracy of the Spp-U-net network. The accuracy of Spp-U-net network algorithm is improved compared with U-net model. The model has strong real-time performance on the road segmentation problem under the premise of ensuring accuracy.

We researched the diagnostic capabilities of deep learning on chest radiographs and an image classifier based on the COVID-Net was presented to classify chest X-Ray images. In the case of a small amount of COVID-19 data, data enhancement was proposed to expanded COVID-19 data 17 times. Our model aims at transfer learning, model integration and classify chest X-Ray images according to three labels: normal, COVID-19 and viral pneumonia. According to the accuracy and loss value, choose the models ResNet-101 and ResNet-152 with good effect for fusion, and dynamically improve their weight ratio during the training process. After training, the model can achieve 96.1% of the types of chest X-Ray images accuracy on the test set. This technology has higher sensitivity than radiologists in the screening and diagnosis of lung nodules. As an auxiliary diagnostic technology, it can help radiologists improve work efficiency and diagnostic accuracy.

Semantic segmentation of an image scene provides semantic information of image regions while less information of objects. In this paper, we propose a method of hierarchical semantic segmentation, including scene level and object level, which aims at labeling both scene regions and objects in an image. In the scene level, we use a feature-based MRF model to recognize the scene categories. The raw probability for each category is predicted via a one-vs-all classification mode. The features and raw probability of superpixels are embedded into the MRF model. With the graph-cut inference, we get the raw scene-level labeling result. In the object level, we use a constraint-based geodesic propagation to get object segmentation. The category and appearance features are utilized as the prior constraints to guide the direction of object label propagation. In this hierarchical model, the scene-level labeling and the object-level labeling have a mutual relationship, which regions and objects are optimized interactively. The experimental results on two datasets show the well performance of our method.

The popularity of social networks has brought the rapid growth of social images which have become an increasingly important image type. One of the most obvious attributes of social images is the tag. However, the sate-of-the-art methods fail to fully exploit the tag information for saliency detection. Thus this paper focuses on salient region detection of social images using both image appearance features and image tag cues. First, a deep convolution neural network is built, which considers both appearance features and tag features. Second, tag neighbor and appearance neighbor based saliency aggregation terms are added to the saliency model to enhance salient regions. The aggregation method is dependent on individual images and considers the performance gaps appropriately. Finally, we also have constructed a new large dataset of challenging social images and pixel-wise saliency annotations to promote further researches and evaluations of visual saliency models. Extensive experiments show that the proposed method performs well on not only the new dataset but also several state-of-the-art saliency datasets.

In the real world, people often focus on the distinctive objects (Salient Regions, SR) in a scene. Thus, a number of saliency detection methods are introduced into content-based image retrieval (CBIR), which is often with Bag of Words (BoW) model. These methods aim to use the saliency map to prune keypoints or discard the keypoints from the background. However, these methods do not consider the background of the image and the characteristics of the dataset itself. In this paper we focus on the following two issues: 1) whether the saliency pruning method is useful for image retrieval in different kinds of datasets (e.g., salient/cluttered, mixed image database); 2) we test the effectiveness of the discarded parts from the background (Non-Salient Regions, Non-SR) for different kinds of image database. In order to demonstrate the performance of using visual saliency, we conduct experiments on two publicly available database (Ukbench, Holidays). The experiments reveal that the way of using saliency map to filter a small amount of key-points can clearly improve the performance of CBIR, and the keypoints in the background are also useful in some kinds of image datasets.

This paper proposes a real-time traffic light detection and recognition algorithm that would allow for the recognition of traffic signals in intelligent vehicles. This algorithm is based on C-HOG features (Color and HOG features) and Support Vector Machine (SVM). The algorithm extracted red and green areas in the video accurately, and then screened the eligible area. Thereafter, the C-HOG features of all kinds of lights could be extracted. Finally, this work used SVM to build a classifier of corresponding category lights. This algorithm obtained accurate real-time information based on the judgment of the decision function. Furthermore, experimental results show that this algorithm demonstrated accuracy and good real-time performance.

As an important learning scheme for Multi-Instance Learning (MIL), the Instance Prototype (IP) selection-based MIL algorithms transform bags into a new instance feature space and achieve impressed classification performance. However, the number of IPs in the existing algorithms linearly increases with the scale of the training data. The performance and efficiencies of these algorithms are easily limited by the high dimension and noise when facing a large scale of training data. This paper proposes a discriminative subspaces-based instance prototype selection method that is suitable for reducing the computation complexity for large scale training data. In the proposed algorithm, we introduce the low-rank matrix recovery technique to find two discriminative and clean subspaces with less noise; then present a l(2,1) norm-based self-expressive sparse coding model to select the most representative instances in each subspace. Experimental results on several data sets show that our algorithm achieves superior and stable performance but with lower dimension compared with other IP selection strategies. (C) 2015 Elsevier B.V. All rights reserved.

In this paper, we propose a novel exposure fusion scheme using the sparse representation theory, which can explore the sparseness of the source images. First, we present a novel way to get the chrominance information of the scene, and the saturation of the fused image can be adjusted using one user-controlled parameter. Second, we conduct the sparse representation on overlapping patches of luminance images obtained by 'sliding window technique', which use dictionary obtained by K-SVD with typical indoor and outdoor multiple exposure sequences. In addition, we introduce an efficient implementation of K-SVD (called approximate K-SVD) which can reduce complexity as well as memory requirements. Third, the coefficients are combined with a novel "frequency of atoms usage" fusion rule strategy. Finally, the fused image is reconstructed from the combined sparse coefficients and the used dictionary. Experiments show that the proposed method can give comparative results compared to state-of-art exposure fusion methods.

Social image tag ranking has emerged as an important research topic due to its application on web image search. This paper presents an adaptive all-season tag ranking algorithm which can handle the images with and without distinct object(s) using different tag ranking strategies. Firstly, based on saliency map derived from the visual attention model, a linear SVM is trained to pre-classify an image as attentive or non-attentive category by using the gray histogram descriptor on the corresponding saliency map. Then, an image with distinct object is processed by the tag saliency ranking algorithm emphasizing distinct object, which combines image saliency map with sparse representation based multi-instance learning algorithm. On the other hand, an image without distinct object can be processed by the tag relevance ranking algorithm via the sparse representation based neighbor-voting strategy. Such adaptive all-season tag ranking strategy can be regarded as taking full advantage of existing tag ranking paradigms. Experiments conducted on well-known image data sets demonstrate the effectiveness of the proposed framework.

In this paper, a sub-dictionary based sparse coding method is proposed for image representation. The novel sparse coding method substitutes a new regularization item for L1-norm in the sparse representation model. The proposed sparse coding method involves a series of sub-dictionaries. Each sub-dictionary contains all the training samples except for those from one particular category. For the test sample to be represented, all the sub-dictionaries should linearly represent it apart from the one that does not contain samples from that label, and this sub-dictionary is called irrelevant sub-dictionary. This new regularization item restricts the sparsity of each sub-dictionary's residual, and this restriction is helpful for classification. The experimental results demonstrate that the proposed method is superior to the previous related sparse representation based classification.

Salient object detection is an important process for machines to understand visual contents as humans. Typically, most previous studies on salient object detection infer salient map by only using the visual features. In this paper, we propose a new paradigm on salient object detection, which aims at producing more reliable results by mining the context information from user's annotated tags. To address this problem, we firstly construct a large scale salient object dataset, which includes 5429 images from the NUSWIDE dataset (a real world web image database from National University of Singapore) with tag information and accurate human-labeled masks. Moreover, a specialised conditional random field (CRF) model is also proposed which takes account of both tag contexts and appearance cues. Extensive experiments show the effectiveness and superiority of our model which can accurately locate the salient objects.

The attribute based vector space model generalizes standard representations of similarity concept in terms of tree architecture. In the model, every concept in the hierarchical tree has its collections of attributes including common and distinctive parts, and the probability of the attributes attached to the concept. A concept is represent as an attribute based vector space, and the similarity is described as feature matching process with cosine similarity measure. The model contains node depth information, node density information of the tree architecture inherent and hidden in it, we show that this measure compares favorably to other measures. This measure is flexible in that it can make comparisons between any two concepts in a hierarchical tree without regard to corpus and dictionary information.

The paper proposes a new spatial extension of Bag-of-Features BoF formalism for classification tasks. The scheme is based on multi-annulus partition which contains much spatial information of image space. Experiments are conducted using final super-vector image representation in Support Vector Machine SVM framework for classification on Oxford flowers and 15 scenes data sets. The results of experiment have shown the effectiveness of our scheme in terms of multiple performance metrics. In addition, our scheme is conceptually simple and easily adoptable. It can lead to much more compact representations and more invariance to image transformation compared to several existing works.