Enhancing Training Collections for Image Annotation: An Instance-Weighted Mixture Modeling Approach

Sawant, Neela; Wang, James; Li, Jia

doi:10.1109/tip.2013.2262289

Cited by 5 publications

(7 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hence, we firstly smooth the loss and then use Nesterov's optimal gradient method [12] to solve (5). According to [12], the smoothed version of the hinge loss g(h k , y k , θ) = max{0, −y k (1 − θ T h k )} can be given by…”

Section: B Optimization Proceduresmentioning

confidence: 99%

“…The loss function Φ(θ) is nondifferentiable. Hence, we firstly smooth the loss and then use Nesterov's optimal gradient method [12] to solve (5). According to [12], the smoothed version of the hinge loss…”

Section: B Optimization Proceduresmentioning

confidence: 99%

“…We adopt the Nesterov's method to solve the smoothed version of problem (5) since it can achieve the optimal convergence rate at O(1/k 2 ), which indicates a low time complexity [12], [34]. To utilize Nesterov's method for optimization, we have to compute the gradient of the smoothed hinge loss to determine the descent direction, as well as the Lipschitz constant to determine the step size of each iteration.…”

Section: B Optimization Proceduresmentioning

confidence: 99%

“…Finally, based on the obtained gradient and Lipschitz constant, we apply Nesterov's method to minimize the smoothed primal F σ (θ). In the t'th iteration round, two auxiliary optimizations are constructed and their solutions are used to build the solution of problem (5). We use θ t , y t and z t to represent the solutions of DTDML w.r.t.…”

Section: B Optimization Proceduresmentioning

confidence: 99%

“…Realworld applications, e.g. image annotation [4], [5], usually have few training samples in the instance space of the target learning task due to the high labeling cost. However, we can easily obtain a large number of labeled samples from the instance spaces of different, but related learning tasks, or from the same instance space with different distribution.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Decomposition-Based Transfer Distance Metric Learning for Image Classification

Luo

Liu

Tao

et al. 2014

IEEE Trans. on Image Process.

107

View full text Add to dashboard Cite

Distance metric learning (DML) is a critical factor for image analysis and pattern recognition. To learn a robust distance metric for a target task, we need abundant side information (i.e., the similarity/dissimilarity pairwise constraints over the labeled data), which is usually unavailable in practice due to the high labeling cost. This paper considers the transfer learning setting by exploiting the large quantity of side information from certain related, but different source tasks to help with target metric learning (with only a little side information). The state-of-the-art metric learning algorithms usually fail in this setting because the data distributions of the source task and target task are often quite different. We address this problem by assuming that the target distance metric lies in the space spanned by the eigenvectors of the source metrics (or other randomly generated bases). The target metric is represented as a combination of the base metrics, which are computed using the decomposed components of the source metrics (or simply a set of random bases); we call the proposed method, decomposition-based transfer DML (DTDML). In particular, DTDML learns a sparse combination of the base metrics to construct the target metric by forcing the target metric to be close to an integration of the source metrics. The main advantage of the proposed method compared with existing transfer metric learning approaches is that we directly learn the base metric coefficients instead of the target metric. To this end, far fewer variables need to be learned. We therefore obtain more reliable solutions given the limited side information and the optimization tends to be faster. Experiments on the popular handwritten image (digit, letter) classification and challenge natural image annotation tasks demonstrate the effectiveness of the proposed method.

show abstract

Section: B Optimization Proceduresmentioning

confidence: 99%

Section: B Optimization Proceduresmentioning

confidence: 99%

Section: B Optimization Proceduresmentioning

confidence: 99%

Section: B Optimization Proceduresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Decomposition-Based Transfer Distance Metric Learning for Image Classification

Luo

Liu

Tao

et al. 2014

IEEE Trans. on Image Process.

107

View full text Add to dashboard Cite

show abstract

Indoor and outdoor image classification

Velswamy

Devadass²,

Velswamy

et al. 2019

IJIUS

View full text Add to dashboard Cite

Purpose The purpose of this paper is to classify the given image as indoor or outdoor with higher success rate by mixing various features like brightness, number of straight lines, number of Euclidean shapes and recursive shapes. Design/methodology/approach For annotating an image, it is very easy, if the image is categorized as indoor or outdoor. Many methods are proposed to classify the given image in these criteria but still the rate of uncategorized images occupies considerable area. This proposed work is the extension of the existing works already proposed by experts in this field. Some of the parameters mainly focused to classify are color histogram, orientation of edges, straightness of edges, discrete cosine transform coefficients, etc. In addition to that, this work includes finding of Euclidean shapes i.e. closed contours and recursive shapes in the given image. When the Euclidean shaped object dominates the recursive shapes then it is classified as indoor object and if the recursive shapes dominates, it is categorized as outdoor object. Findings This work is carried out on the standard image data sets. The data sets are Microsoft Research Cambridge (MRC) object recognition image database 1.0. and Kodak and Coral image data set. Totally 540 images are taken into account and the images are classified 95.4 percent correctly. Originality/value Many methods are proposed to classify the given image in these criteria but still the rate of uncategorized images occupies considerable area. This proposed work is the extension of the existing works already proposed by experts in this field. Some of the parameters mainly focused to classify are color histogram, orientation of edges, straightness of edges, discrete cosine transform coefficients, etc. In addition to that, this work includes finding of Euclidean shapes i.e. closed contours and recursive shapes in the given image. When the Euclidean shaped object dominates the recursive shapes then it is classified as indoor object and if the recursive shapes dominates, it is categorized as outdoor object. This work is carried out on the standard image data sets. The data sets are MRC object recognition image database 1.0. and Kodak and Coral image data set. Totally 540 images are taken into account and the images are classified 95.4 percent correctly.

show abstract

Joint Segmentation and Recognition of Categorized Objects From Noisy Web Image Collection

Wang

Hua

Xue

et al. 2014

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

The segmentation of categorized objects addresses the problem of joint segmentation of a single category of object across a collection of images, where categorized objects are referred to objects in the same category. Most existing methods of segmentation of categorized objects made the assumption that all images in the given image collection contain the target object. In other words, the given image collection is noise free. Therefore, they may not work well when there are some noisy images which are not in the same category, such as those image collections gathered by a text query from modern image search engines. To overcome this limitation, we propose a method for automatic segmentation and recognition of categorized objects from noisy Web image collections. This is achieved by cotraining an automatic object segmentation algorithm that operates directly on a collection of images, and an object category recognition algorithm that identifies which images contain the target object. The object segmentation algorithm is trained on a subset of images from the given image collection which are recognized to contain the target object with high confidence, while training the object category recognition model is guided by the intermediate segmentation results obtained from the object segmentation algorithm. This way, our co-training algorithm automatically identifies the set of true positives in the noisy Web image collection, and simultaneously extracts the target objects from all the identified images. Extensive experiments validated the efficacy of our proposed approach on four datasets: 1) the Weizmann horse dataset, 2) the MSRC object category dataset, 3) the iCoseg dataset, and 4) a new 30-categories dataset including 15,634 Web images with both hand-annotated category labels and ground truth segmentation labels. It is shown that our method compares favorably with the state-of-the-art, and has the ability to deal with noisy image collections.

show abstract

Enhancing Training Collections for Image Annotation: An Instance-Weighted Mixture Modeling Approach

Cited by 5 publications

References 51 publications

Decomposition-Based Transfer Distance Metric Learning for Image Classification

Decomposition-Based Transfer Distance Metric Learning for Image Classification

Indoor and outdoor image classification

Joint Segmentation and Recognition of Categorized Objects From Noisy Web Image Collection

Contact Info

Product

Resources

About