We propose and evaluate several deep network architectures for measuring the similarity between sketches and photographs, within the context of the sketch based image retrieval (SBIR) task. We study the ability of our networks to generalize across diverse object categories from limited training data, and explore in detail strategies for weight sharing, pre-processing, data augmentation and dimensionality reduction. In addition to a detailed comparative study of network configurations, we contribute by describing a hybrid multi-stage training network that exploits both contrastive and triplet networks to exceed state of the art performance on several SBIR benchmarks by a significant margin. Datasets and models are available at www.cvssp.org.
Detecting anomalous activity in video surveillance often involves using only normal activity data in order to learn an accurate detector. Due to lack of annotated data for some specific target domain, one could employ existing data from a source domain to produce better predictions. Hence, transfer learning presents itself as an important tool. But how to analyze the resulting data space? This paper investigates video anomaly detection, in particular feature embeddings of pre-trained CNN that can be used with non-fully supervised data. By proposing novel cross-domain generalization measures, we study how source features can generalize for different target video domains, as well as analyze unsupervised transfer learning. The proposed generalization measures are not only a theorical approach, but show to be useful in practice as a way to understand which datasets can be used or transferred to describe video frames, which it is possible to better discriminate between normal and anomalous activity.
A feature learning task involves training models that are capable of inferring good representations (transformations of the original space) from input data alone. When working with limited or unlabelled data, and also when multiple visual domains are considered, methods that rely on large annotated datasets, such as Convolutional Neural Networks (CNNs), cannot be employed. In this paper we investigate different auto-encoder (AE) architectures, which require no labels, and explore training strategies to learn representations from images. The models are evaluated considering both the reconstruction error of the images and the feature spaces in terms of their discriminative power. We study the role of dense and convolutional layers on the results, as well as the depth and capacity of the networks, since those are shown to affect both the dimensionality reduction and the capability of generalising for different visual domains. Classification results with AE features were as discriminative as pre-trained CNN features. Our findings can be used as guidelines for the design of unsupervised representation learning methods within and across domains.
We present an algorithm for visually searching image collections using free-hand sketched queries. Prior sketch based image retrieval (SBIR) algorithms adopt either a category-level or fine-grain (instancelevel) definition of cross-domain similarity-returning images that match the sketched object class (category-level SBIR), or a specific instance of that object (fine-grain SBIR). In this paper we take the middle-ground; proposing an SBIR algorithm that returns images sharing both the object category and key visual characteristics of the sketched query without assuming photo-approximate sketches from the user. We describe a deeply learned cross-domain embedding in which 'mid-grain' sketch-image similarity may be measured, reporting on the efficacy of unsupervised and semi-supervised manifold alignment techniques to encourage better intracategory (mid-grain) discrimination within that embedding. We propose a new mid-grain sketch-image dataset (MidGrain65c) and demonstrate not only mid-grain discrimination, but also improved category-level discrimination using our approach.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.