multi-instance multi-label learning, an extension of multi-instancelearning in multi-label classification, has been successfully used in image classification. In existing algorithms, the distribution of instances in ...
详细信息
ISBN:
(纸本)9781467376068
multi-instance multi-label learning, an extension of multi-instancelearning in multi-label classification, has been successfully used in image classification. In existing algorithms, the distribution of instances in bags is generally assumed to be independent of each other, which is difficult to be guaranteed in image classification. Considering instance correlations in a bag, in this paper a novel method of scene classification based on multi-kernel fusion and multi-instance multi-label learning is proposed. First, instance correlations are introduced by means of building graph. Then, different kernel matrices can be derived from kernel functions based on graphs in different scales. Finally, the multi-label can be predicted by the multi-kernel SVM classifier based on multiple-kernel fusion. Experimental results on scene data set and MSRC v2 data set show that the proposed method greatly improves the accuracy of the image scene classification compared with other methods.
multi-instance multi-label learning (MIML) is a novel learning framework where each sample is represented by multiple instances and associated with multiple class labels. In several learning situations, the multi-inst...
详细信息
multi-instance multi-label learning (MIML) is a novel learning framework where each sample is represented by multiple instances and associated with multiple class labels. In several learning situations, the multi-instancemulti-label REF neural networks (MIMLRBF) can exploit connections between the instances and the labels of an MIML example directly. However, it is quite often that the numbers of samples in different categories are discrete, i.e., the class distribution is imbalanced. When an MIMLRBF is trained with imbalanced samples, it will produce poor performance for setting the consistent fraction parameter a for all classes. This paper presents an improved approach in learning algorithms used for training MIMLRBF with imbalanced samples. In the first cluster stage, the methodology calculates the initial medoids for each category based on the data density. Afterwards, k-medoids is been invoked to optimize the medoids. The network will take advantage of the well-adjusted units. In the second stage, the weights between the first and second layer are optimized by the singular value decomposition method. The improved approaches could be used in applications with imbalanced samples. Comparing results employing diverse learning strategies shows interesting outcomes as have come out of this paper.
In multi-instance multi-label learning (MIML), each example is not only represented by multiple instances but also associated with multiple class labels. Several learning frameworks, such as the traditional supervised...
详细信息
In multi-instance multi-label learning (MIML), each example is not only represented by multiple instances but also associated with multiple class labels. Several learning frameworks, such as the traditional supervised learning. can be regarded as degenerated versions of MIML. Therefore, an intuitive way to solve MIML problem is to identify its equivalence in its degenerated versions. However, this identification process would make useful information encoded in training examples get lost and thus impair the learning algorithm's performance. In this paper, RBF neural networks are adapted to learn from MIML examples. Connections between instances and labels are directly exploited in the process of first layer clustering and second layer optimization. The proposed method demonstrates superior performance on two real-world MIML tasks. (c) 2009 Elsevier B.V. All rights reserved.
multi-instance multi-label learning (MIML) is an innovative learning framework where each sample is represented by multiple instances and associated with multiple class labels. In several learning situations, the mult...
详细信息
multi-instance multi-label learning (MIML) is an innovative learning framework where each sample is represented by multiple instances and associated with multiple class labels. In several learning situations, the multi-instancemulti-label RBF neural networks (MIMLRBF) can exploit connections between the instances and the labels of an MIML example directly, while most of other algorithms cannot learn that directly. However, the singular value decomposition (SVD) method used to compute the weights of the output layer will cause augmented overall error in network performance when training data are noisy or not easily discernible. This paper presents an improved approach to learning algorithms used for training MIMLRBF. The steepest descent (SD) method is used to optimize the weights after they are initialized by the SVD method. Comparing results employing diverse learning strategies shows interesting outcomes as have come out of this paper.
In this paper, a multi-instancemulti-label algorithm based on neural networks is proposed for image classification. The proposed algorithm, termed multi-instancemulti-label neural network (MIMLNN), consists of two s...
详细信息
In this paper, a multi-instancemulti-label algorithm based on neural networks is proposed for image classification. The proposed algorithm, termed multi-instancemulti-label neural network (MIMLNN), consists of two stages of multiLayer Perceptrons (MLP). For multi-instancemulti-label image classification, all the regional features are fed to the first-stage MLP, with one MLP copy processing one image region. After that, the MLP in the second stage incorporates the outputs of the first-stage MLPs to produce the final labels for the input image. The first-stage MLP is expected to model the relationship between regions and labels, while the second-stage MLP aims at capturing the label correlation for classification refinement. Error Back-Propagation (BP) approach is adopted to tune the parameters of MIMLNN. In view of that traditional gradient descent algorithm suffers from long-term dependency problem, a refined BP algorithm named Rprop is extended to effectively train MIMLNN. The experiments are conducted on a synthetic dataset and the Corel dataset. Experimental results demonstrate the superior performance of MIMLNN comparing with state-of-the-art algorithms for multi-instancemulti-label image classification. (c) 2012 Elsevier B.V. All rights reserved.
Recently, various bag-of-features (BoF) methods show their good resistance to within-class variations and occlusions in object categorization. In this paper, we present a novel approach for multi-object categorization...
详细信息
Recently, various bag-of-features (BoF) methods show their good resistance to within-class variations and occlusions in object categorization. In this paper, we present a novel approach for multi-object categorization within the BoF framework. The approach addresses two issues in BoF related methods simultaneously: how to avoid scene modeling and how to predict labels of an image when multiple categories of objects are co-existing. We employ a biased sampling strategy which combines the bottom-up, biologically inspired saliency information and loose, top-down class prior information for object class modeling. Then this biased sampling component is further integrated with a multi-instancemulti-label leaning and classification algorithm. With the proposed biased sampling strategy, we can perform multi-object categorization within an image without semantic segmentation. The experimental results on PASCAL VOC2007 and SUN09 show that the proposed method significantly improves the discriminative ability of BoF methods and achieves good performance in multi-object categorization tasks.
With the rapid development of digital cameras, we have witnessed great interest and promise in automatic image annotation as a hot research field. Automatic image annotation is an effective method to resolve the probl...
详细信息
With the rapid development of digital cameras, we have witnessed great interest and promise in automatic image annotation as a hot research field. Automatic image annotation is an effective method to resolve the problem of “Semantic Gap”. Automatic Image annotation is a challenging problem when a label is provided for the entire training image only instead of the object region. To eliminate labeling ambiguity, image categorization and object localization should be performed simultaneously. Discriminative multiple instancemulti-labellearning (MIML) can be used for this task by regarding each image as a bag and sub-windows in the image as instances. learning a discriminative multi-instance classifier requires an iterative solution. In each round, positive sub-windows for the next round should be selected. With standard approaches, selecting only one positive sub-window per positive bag may limit the search space for global optimum; meanwhile, selecting all temporal positive sub-windows may add noise into learning. We select a subset of sub-windows per positive bag to avoid those limitations. Our Proposed EMID algorithm is able to take the correlation among instances, correlation among labels, and correlation between instances and labels simultaneously, and provides a very rich representation and learning potential. Experimental results demonstrate that our approach outperforms previous discriminative MIML approaches and standard categorization approaches.
暂无评论