scispace - formally typeset
Search or ask a question
Book ChapterDOI

Image annotation using metric learning in semantic neighbourhoods

TL;DR: 2PKNN, a two-step variant of the classical K-nearest neighbour algorithm, is proposed that performs comparable to the current state-of-the-art on three challenging image annotation datasets, and shows significant improvements after metric learning.
Abstract: Automatic image annotation aims at predicting a set of textual labels for an image that describe its semantics. These are usually taken from an annotation vocabulary of few hundred labels. Because of the large vocabulary, there is a high variance in the number of images corresponding to different labels ("class-imbalance"). Additionally, due to the limitations of manual annotation, a significant number of available images are not annotated with all the relevant labels ("weak-labelling"). These two issues badly affect the performance of most of the existing image annotation models. In this work, we propose 2PKNN, a two-step variant of the classical K-nearest neighbour algorithm, that addresses these two issues in the image annotation task. The first step of 2PKNN uses "image-to-label" similarities, while the second step uses "image-to-image" similarities; thus combining the benefits of both. Since the performance of nearest-neighbour based methods greatly depends on how features are compared, we also propose a metric learning framework over 2PKNN that learns weights for multiple features as well as distances together. This is done in a large margin set-up by generalizing a well-known (single-label) classification metric learning algorithm for multi-label prediction. For scalability, we implement it by alternating between stochastic sub-gradient descent and projection steps. Extensive experiments demonstrate that, though conceptually simple, 2PKNN alone performs comparable to the current state-of-the-art on three challenging image annotation datasets, and shows significant improvements after metric learning.

Content maybe subject to copyright    Report

Citations
More filters
Proceedings ArticleDOI
01 Apr 2014
TL;DR: A sparse kernel learning framework for the Continuous Relevance Model (CRM) that greedily selects an optimal combination of kernels, which rapidly converges to an annotation accuracy that substantially outperforms a host of state-of-the-art annotation models.
Abstract: In this paper we introduce a sparse kernel learning framework for the Continuous Relevance Model (CRM). State-of-the-art image annotation models linearly combine evidence from several different feature types to improve image annotation accuracy. While previous authors have focused on learning the linear combination weights for these features, there has been no work examining the optimal combination of kernels. We address this gap by formulating a sparse kernel learning framework for the CRM, dubbed the SKL-CRM, that greedily selects an optimal combination of kernels. Our kernel learning framework rapidly converges to an annotation accuracy that substantially outperforms a host of state-of-the-art annotation models. We make two surprising conclusions: firstly, if the kernels are chosen correctly, only a very small number of features are required so to achieve superior performance over models that utilise a full suite of feature types; and secondly, the standard default selection of kernels commonly used in the literature is sub-optimal, and it is much better to adapt the kernel choice based on the feature type and image dataset.

33 citations


Cites background from "Image annotation using metric learn..."

  • ...The CRM estimates the joint probability distribution of a set of words w = {w1 . . . wK} from a vocabulary of size V together with an image f represented as a set of feature vectors f = {~f1. . . ~fM}....

    [...]

  • ...…required so to achieve superior performance over models that utilise a full suite of feature types; and secondly, the standard default selection of kernels commonly used in the literature is sub-optimal, and it is much better to adapt the kernel choice based on the feature type and image dataset....

    [...]

Journal ArticleDOI
TL;DR: This paper introduces a novel approach based on non-linear matrix completion for image tagging task with defective tags and presents a formal methodology together with an optimization method under the matrix completion framework to jointly complete the tags of training and test images.
Abstract: In this paper, we address the problem of social image tagging using practical vocabulary for mobile users on the social media. On the social media, images usually have an incomplete or noisy set of social tags provided by the mobile users, and we consider this issue as defective tag assignments . Previous studies on social image tagging have mostly focused on multi-label classification without considering the defective tags. In these studies, the usage of multi-label classification techniques is expected to synergically exploit the linear relations between the image features and the semantic tags. However, these approaches usually aimed to capture the linear relations from the training data while ignoring the helpful information from the test data. In addition, they failed to incorporate the non-linear associations residing in the visual features as well as in the semantic tags. To overcome these drawbacks, we introduce a novel approach based on non-linear matrix completion for image tagging task with defective tags. Specifically, we first construct the entire feature-tag matrix based on the visual features with non-linear kernel mapping. Then, we present a formal methodology together with an optimization method under the matrix completion framework to jointly complete the tags of training and test images. Experimental evaluations demonstrate that our method shows promising results on image tagging task on two benchmark social image datasets with defective tags, and establishes a baseline for such models in this research domain.

32 citations


Cites background from "Image annotation using metric learn..."

  • ...Generally, previous researches for image annotation can be roughly categorized into three groups: generative models [25]–[27], discriminative models [23], [28] and nearest neighbor (NN) based models [10]–[12], [29], [30]....

    [...]

Journal ArticleDOI
TL;DR: A new form of the continuous relevance model (CRM), dubbed the SKL-CRM, that adaptively selects the best performing kernel per feature type for automatic image annotation is introduced and is found to attain performance that is competitive to a suite of state-of-the-art image annotation models.
Abstract: In this paper, we introduce a new form of the continuous relevance model (CRM), dubbed the SKL-CRM, that adaptively selects the best performing kernel per feature type for automatic image annotation. Previous image annotation models apply a standard selection of kernels to model the distribution of image features. Popular examples include a Gaussian kernel for modelling GIST features or a Laplacian kernel for global colour histograms. In this work, we demonstrate that this standard assignment of kernels to feature types is sub-optimal and a substantially higher image annotation accuracy can be attained by adapting the kernel-feature assignment. We formulate an efficient greedy algorithm to find the best kernel-feature alignment and show that it is able to rapidly find a sparse subset of features that maximises annotation $$F_{1}$$ score. In a second contribution, we introduce two data-adaptive kernels for image annotation—the generalised Gaussian and multinomial kernels—which we demonstrate can better model the distribution of image features as compared to standard kernels. Evaluation is conducted on three standard image datasets across a selection of different feature representations. The proposed SKL-CRM model is found to attain performance that is competitive to a suite of state-of-the-art image annotation models.

26 citations


Cites background or methods or result from "Image annotation using metric learn..."

  • ...HYP-2 Learning an optimal combination of kernels using the data itself, owing to its different geometry over the feature space, will outperform the standard (default) assignment of kernels to feature types often found in the literature [17,45]....

    [...]

  • ...The training and testing dataset splits for all three datasets are identical to previously published work [17,45]....

    [...]

  • ...All datasets are identical to those used in most recent image annotation publications [17,45], thereby permitting direct comparison....

    [...]

  • ...We use the identical subset of images as [45]....

    [...]

  • ...To fairly compare our model performance to previously published figures we use the identical feature set, parameter optimisation strategy and evaluation procedure of previous relevant work [17,26,45]....

    [...]

Journal ArticleDOI
TL;DR: An AIA system using Non-negative Matrix Factorization (NMF) framework, which discovers a latent space, by factorizing data into a set of non-negative basis and coefficients and is competitive with the current state-of-the-art methods.

25 citations


Cites background or methods from "Image annotation using metric learn..."

  • ...But, for a fair comparison with other methods, we used the same feature set as in [20] and [21], without any claim of being the best feature set....

    [...]

  • ...0 238 - - - 2PKNN+ML [21] 44 46 45 191 53 27 35....

    [...]

  • ...The previous works [12, 20, 21], simply chose a fixed number of tags (5 tags) to annotate all test images....

    [...]

  • ...To the best of our knowledge, the best search-based work until now is 2PKNN [21] which uses a two-level multi-label metric learning to learn the weights of each feature and also the weight of each element of a single feature vector....

    [...]

Journal ArticleDOI
TL;DR: A method for automatic video annotation that increases the number of tags originally provided by users, and localizes them temporally, associating tags to keyframes, which exploits collective knowledge embedded in user-generated tags and web sources.

25 citations


Cites background from "Image annotation using metric learn..."

  • ...much attention for image annotation and tag relevance estimation [3, 4, 5]....

    [...]

References
More filters
Proceedings Article
05 Dec 2005
TL;DR: In this article, a Mahanalobis distance metric for k-NN classification is trained with the goal that the k-nearest neighbors always belong to the same class while examples from different classes are separated by a large margin.
Abstract: We show how to learn a Mahanalobis distance metric for k-nearest neighbor (kNN) classification by semidefinite programming. The metric is trained with the goal that the k-nearest neighbors always belong to the same class while examples from different classes are separated by a large margin. On seven data sets of varying size and difficulty, we find that metrics trained in this way lead to significant improvements in kNN classification—for example, achieving a test error rate of 1.3% on the MNIST handwritten digits. As in support vector machines (SVMs), the learning problem reduces to a convex optimization based on the hinge loss. Unlike learning in SVMs, however, our framework requires no modification or extension for problems in multiway (as opposed to binary) classification.

4,433 citations

Journal ArticleDOI
TL;DR: This paper shows how to learn a Mahalanobis distance metric for kNN classification from labeled examples in a globally integrated manner and finds that metrics trained in this way lead to significant improvements in kNN Classification.
Abstract: The accuracy of k-nearest neighbor (kNN) classification depends significantly on the metric used to compute distances between different examples. In this paper, we show how to learn a Mahalanobis distance metric for kNN classification from labeled examples. The Mahalanobis metric can equivalently be viewed as a global linear transformation of the input space that precedes kNN classification using Euclidean distances. In our approach, the metric is trained with the goal that the k-nearest neighbors always belong to the same class while examples from different classes are separated by a large margin. As in support vector machines (SVMs), the margin criterion leads to a convex optimization based on the hinge loss. Unlike learning in SVMs, however, our approach requires no modification or extension for problems in multiway (as opposed to binary) classification. In our framework, the Mahalanobis distance metric is obtained as the solution to a semidefinite program. On several data sets of varying size and difficulty, we find that metrics trained in this way lead to significant improvements in kNN classification. Sometimes these results can be further improved by clustering the training examples and learning an individual metric within each cluster. We show how to learn and combine these local metrics in a globally integrated manner.

4,157 citations


"Image annotation using metric learn..." refers background or methods in this paper

  • ...With this goal, we perform metric learning over 2PKNN by generalizing the LMNN [11] algorithm for multi-label prediction....

    [...]

  • ...In such a scenario, (i) since each base distance contributes differently, we can learn appropriate weights to combine them in the distance space [2, 3]; and (ii) since every feature (such as SIFT or colour histogram) itself is represented as a multidimensional vector, its individual elements can also be weighted in the feature space [11]....

    [...]

  • ...Our extension of LMNN conceptually differs from its previous extensions such as [21] in at least two significant ways: (i) we adapt LMNN in its choice of target/impostors to learn metrics for multi-label prediction problems, whereas [21] uses the same definition of target/impostors as in LMNN to address classification problem in multi-task setting, and (ii) in our formulation, the amount of push applied on an impostor varies depending on its conceptual similarity w.r.t. a given sample, which makes it suitable for multi-label prediction tasks....

    [...]

  • ...Our metric learning framework extends LMNN in two major ways: (i) LMNN is meant for single-label classification (or simply classification) problems, while we adapt it for images annotation which is a multi-label classification task; and (ii) LMNN learns a single Mahalanobis metric in the feature space, while we extend it to learn linear metrics for multi- Image Annotation Using Metric Learning in Semantic Neighbourhoods 3 ple features as well as distances together....

    [...]

  • ...For this purpose, we extend the classical LMNN [11] algorithm for multi-label prediction....

    [...]

Proceedings ArticleDOI
25 Apr 2004
TL;DR: A new interactive system: a game that is fun and can be used to create valuable output that addresses the image-labeling problem and encourages people to do the work by taking advantage of their desire to be entertained.
Abstract: We introduce a new interactive system: a game that is fun and can be used to create valuable output. When people play the game they help determine the contents of images by providing meaningful labels for them. If the game is played as much as popular online games, we estimate that most images on the Web can be labeled in a few months. Having proper labels associated with each image on the Web would allow for more accurate image search, improve the accessibility of sites (by providing descriptions of images to visually impaired individuals), and help users block inappropriate images. Our system makes a significant contribution because of its valuable output and because of the way it addresses the image-labeling problem. Rather than using computer vision techniques, which don't work well enough, we encourage people to do the work by taking advantage of their desire to be entertained.

2,365 citations


"Image annotation using metric learn..." refers background in this paper

  • ...ESP Game contains images annotated using an on-line game, where two (mutually unknown) players are randomly given an image for which they have to predict same keyword(s) to score points [22]....

    [...]

Journal ArticleDOI
TL;DR: A simple and effective stochastic sub-gradient descent algorithm for solving the optimization problem cast by Support Vector Machines, which is particularly well suited for large text classification problems, and demonstrates an order-of-magnitude speedup over previous SVM learning methods.
Abstract: We describe and analyze a simple and effective stochastic sub-gradient descent algorithm for solving the optimization problem cast by Support Vector Machines (SVM). We prove that the number of iterations required to obtain a solution of accuracy $${\epsilon}$$ is $${\tilde{O}(1 / \epsilon)}$$, where each iteration operates on a single training example. In contrast, previous analyses of stochastic gradient descent methods for SVMs require $${\Omega(1 / \epsilon^2)}$$ iterations. As in previously devised SVM solvers, the number of iterations also scales linearly with 1/λ, where λ is the regularization parameter of SVM. For a linear kernel, the total run-time of our method is $${\tilde{O}(d/(\lambda \epsilon))}$$, where d is a bound on the number of non-zero features in each example. Since the run-time does not depend directly on the size of the training set, the resulting algorithm is especially suited for learning from large datasets. Our approach also extends to non-linear kernels while working solely on the primal objective function, though in this case the runtime does depend linearly on the training set size. Our algorithm is particularly well suited for large text classification problems, where we demonstrate an order-of-magnitude speedup over previous SVM learning methods.

2,037 citations


"Image annotation using metric learn..." refers methods in this paper

  • ...To overcome this issue, we solve it by alternatively using stochastic sub-gradient descent and projection steps (similar to Pegasos [12])....

    [...]

  • ...To address this, we implement metric learning by alternating between stochastic sub-gradient descent and projection steps (similar to Pegasos [12])....

    [...]

Book ChapterDOI
28 May 2002
TL;DR: This work shows how to cluster words that individually are difficult to predict into clusters that can be predicted well, and cannot predict the distinction between train and locomotive using the current set of features, but can predict the underlying concept.
Abstract: We describe a model of object recognition as machine translation. In this model, recognition is a process of annotating image regions with words. Firstly, images are segmented into regions, which are classified into region types using a variety of features. A mapping between region types and keywords supplied with the images, is then learned, using a method based around EM. This process is analogous with learning a lexicon from an aligned bitext. For the implementation we describe, these words are nouns taken from a large vocabulary. On a large test set, the method can predict numerous words with high accuracy. Simple methods identify words that cannot be predicted well. We show how to cluster words that individually are difficult to predict into clusters that can be predicted well -- for example, we cannot predict the distinction between train and locomotive using the current set of features, but we can predict the underlying concept. The method is trained on a substantial collection of images. Extensive experimental results illustrate the strengths and weaknesses of the approach.

1,765 citations


"Image annotation using metric learn..." refers background in this paper

  • ...translation models [13, 14] and nearest-neighbour based relevance models [1, 8]....

    [...]

  • ...Corel 5K was first used in [14], and since then it has become a benchmark for comparing annotation performance....

    [...]