Thomas Mensink

Improving the fisher kernel for large-scale image classification

The Fisher kernel (FK) is a generic framework which combines the benefits of generative and discriminative approaches. In the context of image classification the FK was shown to extend the popular bag-of-visual-words (BOV) by going beyond count statistics. However, in practice, this enriched representation has not yet shown its superiority over the BOV. In the first part we show that with several well-motivated modifications over the original framework we can boost the accuracy of the FK. On PASCAL VOC 2007 we increase the Average Precision (AP) from 47.9% to 58.3%. Similarly, we demonstrate state-of-the-art accuracy on CalTech 256. A major advantage is that these results are obtained using only SIFT descriptors and costless linear classifiers. Equipped with this representation, we can now explore image classification on a larger scale. In the second part, as an application, we compare two abundant resources of labeled images to learn classifiers: ImageNet and Flickr groups. In an evaluation involving hundreds of thousands of training images we show that classifiers learned on Flickr groups perform surprisingly well (although they were not intended for this purpose) and that they can complement classifiers learned on more carefully annotated datasets.

Image Classification with the Fisher Vector: Theory and Practice

A standard approach to describe an image for classification and retrieval purposes is to extract a set of local patch descriptors, encode them into a high dimensional vector and pool them into an image-level signature. The most common patch encoding strategy consists in quantizing the local descriptors into a finite set of prototypical elements. This leads to the popular Bag-of-Visual words representation. In this work, we propose to use the Fisher Kernel framework as an alternative patch encoding strategy: we describe patches by their deviation from an “universal” generative Gaussian mixture model. This representation, which we call Fisher vector has many advantages: it is efficient to compute, it leads to excellent results even with efficient linear classifiers, and it can be compressed with a minimal loss of accuracy using product quantization. We report experimental results on five standard datasets—PASCAL VOC 2007, Caltech 256, SUN 397, ILSVRC 2010 and ImageNet10K—with up to 9M images and 10K classes, showing that the FV framework is a state-of-the-art patch encoding technique.

TagProp: Discriminative metric learning in nearest neighbor models for image auto-annotation

Image auto-annotation is an important open problem in computer vision. For this task we propose TagProp, a discriminatively trained nearest neighbor model. Tags of test images are predicted using a weighted nearest-neighbor model to exploit labeled training images. Neighbor weights are based on neighbor rank or distance. TagProp allows the integration of metric learning by directly maximizing the log-likelihood of the tag predictions in the training set. In this manner, we can optimally combine a collection of image similarity metrics that cover different aspects of image content, such as local shape descriptors, or global color histograms. We also introduce a word specific sigmoidal modulation of the weighted neighbor tag predictions to boost the recall of rare words. We investigate the performance of different variants of our model and compare to existing work. We present experimental results for three challenging data sets. On all three, TagProp makes a marked improvement as compared to the current state-of-the-art.

Metric learning for large scale image classification: generalizing to new classes at near-zero cost

We are interested in large-scale image classification and especially in the setting where images corresponding to new or existing classes are continuously added to the training set. Our goal is to devise classifiers which can incorporate such images and classes on-the-fly at (near) zero cost. We cast this problem into one of learning a metric which is shared across all classes and explore k-nearest neighbor (k-NN) and nearest class mean (NCM) classifiers. We learn metrics on the ImageNet 2010 challenge data set, which contains more than 1.2M training images of 1K classes. Surprisingly, the NCM classifier compares favorably to the more flexible k-NN classifier, and has comparable performance to linear SVMs. We also study the generalization performance, among others by using the learned metric on the ImageNet-10K dataset, and we obtain competitive performance. Finally, we explore zero-shot classification, and show how the zero-shot model can be combined very effectively with small training datasets.

Distance-Based Image Classification: Generalizing to New Classes at Near-Zero Cost

We study large-scale image classification methods that can incorporate new classes and training images continuously over time at negligible cost. To this end, we consider two distance-based classifiers, the k-nearest neighbor (k-NN) and nearest class mean (NCM) classifiers, and introduce a new metric learning approach for the latter. We also introduce an extension of the NCM classifier to allow for richer class representations. Experiments on the ImageNet 2010 challenge dataset, which contains over 106 training images of 1,000 classes, show that, surprisingly, the NCM classifier compares favorably to the more flexible k-NN classifier. Moreover, the NCM performance is comparable to that of linear SVMs which obtain current state-of-the-art performance. Experimentally, we study the generalization performance to classes that were not used to learn the metrics. Using a metric learned on 1,000 classes, we show results for the ImageNet-10K dataset which contains 10,000 classes, and obtain performance that is competitive with the current state-of-the-art while being orders of magnitude faster. Furthermore, we show how a zero-shot class prior based on the ImageNet hierarchy can improve performance when few training images are available.

Image annotation with tagprop on the MIRFLICKR set

Image annotation is an important computer vision problem where the goal is to determine the relevance of annotation terms for images. Image annotation has two main applications: (i) proposing a list of relevant terms to users that want to assign indexing terms to images, and (ii) supporting keyword based search for images without indexing terms, using the relevance estimates to rank images. In this paper we present TagProp, a weighted nearest neighbour model that predicts the term relevance of images by taking a weighted sum of the annotations of the visually most similar images in an annotated training set. TagProp can use a collection of distance measures capturing different aspects of image content, such as local shape descriptors, and global colour histograms. It automatically finds the optimal combination of distances to define the visual neighbours of images that are most useful for annotation prediction. TagProp compensates for the varying frequencies of annotation terms using a term-specific sigmoid to scale the weighted nearest neighbour tag predictions. We evaluate different variants of TagProp with experiments on the MIR Flickr set, and compare with an approach that learns a separate SVM classifier for each annotation term. We also consider using Flickr tags to train our models, both as additional features and as training labels. We find the SVMs to work better when learning from the manual annotations, but TagProp to work better when learning from the Flickr tags. We also find that using the Flickr tags as a feature can significantly improve the performance of SVMs learned from manual annotations.

COSTA: Co-Occurrence Statistics for Zero-Shot Classification

In this paper we aim for zero-shot classification, that is visual recognition of an unseen class by using knowledge transfer from known classes. Our main contribution is COSTA, which exploits co-occurrences of visual concepts in images for knowledge transfer. These inter-dependencies arise naturally between concepts, and are easy to obtain from existing annotations or web-search hit counts. We estimate a classifier for a new label, as a weighted combination of related classes, using the co-occurrences to define the weight. We propose various metrics to leverage these co-occurrences, and a regression model for learning a weight for each related class. We also show that our zero-shot classifiers can serve as priors for few-shot learning. Experiments on three multi-labeled datasets reveal that our proposed zero-shot methods, are approaching and occasionally outperforming fully supervised SVMs. We conclude that co-occurrence statistics suffice for zero-shot classification.

Automatic face naming with caption-based supervision

We consider two scenarios of naming people in databases of news photos with captions: (i) finding faces of a single person, and (ii) assigning names to all faces. We combine an initial text-based step, that restricts the name assigned to a face to the set of names appearing in the caption, with a second step that analyzes visual features of faces. By searching for groups of highly similar faces that can be associated with a name, the results of purely text-based search can be greatly ameliorated. We improve a recent graph-based approach, in which nodes correspond to faces and edges connect highly similar faces. We introduce constraints when optimizing the objective function, and propose improvements in the low-level methods used to construct the graphs. Furthermore, we generalize the graph-based approach to face naming in the full data set. In this multi-person naming case the optimization quickly becomes computationally demanding, and we present an important speed-up using graph-flows to compute the optimal name assignments in documents. Generative models have previously been proposed to solve the multi-person naming task. We compare the generative and graph-based methods in both scenarios, and find significantly better performance using the graph-based methods in both cases.

Learning structured prediction models for interactive image labeling

We propose structured models for image labeling that take into account the dependencies among the image labels explicitly. These models are more expressive than independent label predictors, and lead to more accurate predictions. While the improvement is modest for fully-automatic image labeling, the gain is significant in an interactive scenario where a user provides the value of some of the image labels. Such an interactive scenario offers an interesting trade-off between accuracy and manual labeling effort. The structured models are used to decide which labels should be set by the user, and transfer the user input to more accurate predictions on other image labels. We also apply our models to attribute-based image classification, where attribute predictions of a test image are mapped to class probabilities by means of a given attribute-class mapping. In this case the structured models are built at the attribute level. We also consider an interactive system where the system asks a user to set some of the attribute values in order to maximally improve class prediction performance. Experimental results on three publicly available benchmark data sets show that in all scenarios our structured models lead to more accurate predictions, and leverage user input much more effectively than state-of-the-art independent models.

VideoStory: A New Multimedia Embedding for Few-Example Recognition and Translation of Events

This paper proposes a new video representation for few-example event recognition and translation. Different from existing representations, which rely on either low-level features, or pre-specified attributes, we propose to learn an embedding from videos and their descriptions. In our embedding, which we call VideoStory, correlated term labels are combined if their combination improves the video classifier prediction. Our proposed algorithm prevents the combination of correlated terms which are visually dissimilar by optimizing a joint-objective balancing descriptiveness and predictability. The algorithm learns from textual descriptions of video content, which we obtain for free from the web by a simple spidering procedure. We use our VideoStory representation for few-example recognition of events on more than 65K challenging web videos from the NIST TRECVID event detection task and the Columbia Consumer Video collection. Our experiments establish that i) VideoStory outperforms an embedding without joint-objective and alternatives without any embedding, ii) The varying quality of input video descriptions from the web is compensated by harvesting more data, iii) VideoStory sets a new state-of-the-art for few-example event recognition, outperforming very recent attribute and low-level motion encodings. What is more, VideoStory translates a previously unseen video to its most likely description from visual content only.