Learning sparse latent representation and distance metric for image retrieval

The performance of image retrieval depends critically on the semantic representation and the distance function used to estimate the similarity of two images. A good representation should integrate multiple visual and textual (e.g., tag) features and offer a step closer to the true semantics of interest (e.g., concepts). As the distance function operates on the representation, they are interdependent, and thus should be addressed at the same time. We propose a probabilistic solution to learn both the representation from multiple feature types and modalities and the distance metric from data. The learning is regularised so that the learned representation and information-theoretic metric will (i) preserve the regularities of the visual/textual spaces, (ii) enhance structured sparsity, (iii) encourage small intra-concept distances, and (iv) keep inter-concept images separated. We demonstrate the capacity of our method on the NUS-WIDE data. For the well-studied 13 animal subset, our method outperforms state-of-the-art rivals. On the subset of single-concept images, we gain 79:5% improvement over the standard nearest neighbours approach on the MAP score, and 45.7% on the NDCG.

On film character retrieval in feature-length films

The problem of automatic face recognition (AFR) concerns matching a detected (roughly localized) face against a database of known faces with associated identities. This task, although very intuitive to humans and despite the vast amounts of research behind it, still poses a significant challenge to computer-based methods. For reviews of the literature and commercial state-of-the-art see [21, 372] and [252, 253]. Much AFR research has concentrated on the user authentication paradigm (e.g. [10, 30, 183]). In contrast, we consider the content-based multimedia retrieval setup: our aim is to retrieve, and rank by confidence, film shots based on the presence of specific actors. A query to the system consists of the user choosing the person of interest in one or more keyframes.

Automatic face recognition for film character retrieval in feature-length films

The objective of this work is to recognize all the frontal faces of a character in the closed world of a movie or situation comedy, given a small number of query faces. This is challenging because faces in a feature-length film are relatively uncontrolled with a wide variability of scale, pose, illumination, and expressions, and also may be partially occluded. We develop a recognition method based on a cascade of processing steps that normalize for the effects of the changing imaging environment. In particular there are three areas of novelty: (i) we suppress the background surrounding the face, enabling the maximum area of the face to be retained for recognition rather than a subset; (ii) we include a pose refinement step to optimize the registration between the test image and face exemplar; and (iii) we use robust distance to a sub-space to allow for partial occlusion and expression change. The method is applied and evaluated on several feature length films. It is demonstrated that high recall rates (over 92%) can be achieved whilst maintaining good precision (over 93%).

Reading ancient coins: automatically identifying denarii using obverse legend seeded retrieval

The aim of this paper is to automatically identify a Roman Imperial denarius from a single query photograph of its obverse and reverse. Such functionality has the potential to contribute greatly to various national schemes which encourage laymen to report their finds to local museums. Our work introduces a series of novelties: (i) this is the first paper which describes a method for extracting the legend of an ancient coin from a photograph; (ii) we are also the first to suggest the idea and propose a method for identifying a coin using a series of carefully engineered retrievals, each harnessed for further information using visual or meta-data processing; (iii) we show how in addition to a unique standard reference number for a query coin, the proposed system can be used to extract salient coin information (issuing authority, obverse and reverse descriptions, mint date) and retrieve images of other coins of the same type.

Medical image retrieval based on unclean image bags

Traditional content-based image retrieval (CBIR) scheme with assumption of independent individual images in large-scale collections suffers from poor retrieval performance. In medical applications, images usually exist in the form of image bags and each bag includes multiple relevant images of the same perceptual meaning. In this paper, based on these natural image bags, we explore a new scheme to improve the performance of medical image retrieval. It is feasible and efficient to search the bag-based medical image collection by providing a query bag. However, there is a critical problem of noisy images which may present in image bags and severely affect the retrieval performance. A new three-stage solution is proposed to perform the retrieval and handle the noisy images. In stage 1, in order to alleviate the influence of noisy images, we associate each image in the image bags with a relevance degree. In stage 2, a novel similarity aggregation method is proposed to incorporate image relevance and feature importance into the similarity computation process. In stage 3, we obtain the final image relevance in an adaptive way which can consider both image bag similarity and individual image similarity. The experiments demonstrate that the proposed approach can improve the image retrieval performance significantly.

Distance from shore positively influences alert distance in three wetland bird species

Behavioural responses of wetland fauna to humans constitutes a potential conservation threat, and may alter how animals use wetlands and their surrounds. We predicted that the farther from refuge (i.e. water) that terrestrially foraging rails occurred, the longer the distance at which they would become alert when approached. We found that the distance at which Eurasian Coot, Fulica atra, Purple Swamphen, Porphyrio porphyrio, and Dusky Moorhen, Gallinula tenebrosa became alert to an approaching predator (i.e. human) increased with distance from the shore of a wetland (species and starting distance were not significant but associated with low power). Thus, these behavioural measures suggest these birds use water in wetlands as a refuge from potential threats such as people. © 2014 Springer Science+Business Media Dordrecht.

Are vehicles 'mobile bird hides'?: A test of the hypothesis that 'cars cause less disturbance'

We tested the 'cars cause less disturbance' hypothesis by comparing the flight-initiation distance (FID) evoked by a car versus a single walker for 38 species of waterbird (n = 657 standardised approaches). For the 15 species where we had sample size adequate for statistical testing (n ≥ 5), we found that cars elicited shorter responses after controlling for starting distance. Within-species analyses revealed that this difference was significant in 8 of 15 species. Although mean FIDs for car approaches were always shorter than FIDs toward single walkers in the remaining species (7), the tests in those species lacked sufficient power to draw meaningful conclusions. Our results provide support for the hypothesis that birds respond to cars at shorter distances. The wide taxonomic breadth of species investigated suggests that this principle may be broadly applicable, at least in waterbirds. The results of this study and the FID estimates we present will allow development of meaningful stimulus-specific buffer zones to protect waterbirds from disturbance.

Geolocator studies on Ruddy Turnstones Arenaria interpres and Greater Sandplovers Charadrius leschenaultii in the East Asian-Australasia Flyway reveal widely different migration strategies

In 2010, following successful trials with geolocators on Ruddy Turnstones in 2009, a total of 105 units, of four different models, were deployed at five locations on Ruddy Turnstones and Greater Sandplovers. Geolocator retrieval rates were 44% on Ruddy Turnstone and 27% on Greater Sandplover. Complete (59%) and partial (15%) technical failure rates on geolocators were high and were mostly the result of wear and saltwater corrosion. All 30 units from the Swiss Ornithological Institute failed. Only half of the Mk10 and Mk12 units from the British Antarctic Survey produced full migration histories. The northward migration of Ruddy Turnstones was on a narrow path with many birds completing an initial non-stop flight of 7,600 km to Taiwan. Later, most made a stopover in the Yellow Sea. Median migration duration was 39.5 days and average migration speed of the first major leg of the journey (assuming the birds followed the great circle route between stopovers) was 63.4 kph. Southward migration paths showed a much wider spread, ranging from Mongolia to the central Pacific. The latter involved the same bird that had been tracked along this route the previous year. It has now been logged on similar 27,000 km round trips in two successive years. The median duration of southward migration (78 days) was nearly twice that of northward migration and data on average migration speed for just two migration legs indicated that it might be lower, 30 and 40 kph being the values recorded. Greater Sandplovers were only tracked on northward migration but seemed to follow a similar migration strategy with a large initial non-stop flight followed by shorter flights and more regular stopovers. Plans are outlined for further analyses and future deployments of geolocators.

Lack of seasonal and moult-related stress modulation in an opportunistically breeding bird: the white-plumed honeyeater (Lichenostomus penicillatus).

In most vertebrate species, glucocorticoid levels and stress sensitivity vary in relation to season and life-history stage. In birds, baseline corticosterone (CORT) and stress sensitivity are typically highest while breeding and decrease substantially during moult. Because elevated CORT adversely affects protein synthesis, moult-related CORT suppression is thought to be necessary for forming high-quality feathers. Surprisingly, some passerine species lack moult-related CORT suppression, but these are distinguished by having slow rates of moult and being opportunistic breeders. We examined baseline and stress-induced CORT levels in an opportunistically breeding Australian passerine, the white-plumed honeyeater (Lichenostomus penicillatus). Although this species has a slower moult rate than high-latitiude breeders, it differs little from north-temperate passerines. Neither baseline nor stress-induced CORT levels varied with season (winter, spring or summer), sex or moult status in adult birds. While breeding tended to be highest in early spring through late summer, laparotomies revealed only limited reduction in testicular size in males the year round. In all but one sampling period, at least some females displayed follicular hierarchy. Breeding usually coincides with outbreaks of phytophagous insects, which can happen at any time of the year. This results in moult/breeding overlap when infestations occur in late spring or summer. The ability of this species to moult and breed at the same time while having breeding-levels of CORT demonstrates that CORT suppression is not a prerequisite for synthesis of high-quality feathers. An experimental design incorporating moulting and non-moulting phenotypes is suggested to test the functional significance of CORT suppression in other species.

Evaluating non-deterministic retrieval systems

The use of sampling, randomized algorithms, or training based on the unpredictable inputs of users in Information Retrieval often leads to non-deterministic outputs. Evaluating the effectiveness of systems incorporating these methods can be challenging since each run may produce different effectiveness scores. Current IR evaluation techniques do not address this problem. Using the context of distributed information retrieval as a case study for our investigation, we propose a solution based on multivariate linear modeling. We show that the approach provides a consistent and reliable method to compare the effectiveness of non-deterministic IR algorithms, and explain how statistics can safely be used to show that two IR algorithms have equivalent effectiveness. Copyright 2014 ACM.

A clustering-based failure mode and effect analysis model and its application to the edible bird nest industry

Failure mode and effect analysis (FMEA) is a popular safety and reliability analysis tool in examining potential failures of products, process, designs, or services, in a wide range of industries. While FMEA is a popular tool, the limitations of the traditional Risk Priority Number (RPN) model in FMEA have been highlighted in the literature. Even though many alternatives to the traditional RPN model have been proposed, there are not many investigations on the use of clustering techniques in FMEA. The main aim of this paper was to examine the use of a new Euclidean distance-based similarity measure and an incremental-learning clustering model, i.e., fuzzy adaptive resonance theory neural network, for similarity analysis and clustering of failure modes in FMEA; therefore, allowing the failure modes to be analyzed, visualized, and clustered. In this paper, the concept of a risk interval encompassing a group of failure modes is investigated. Besides that, a new approach to analyze risk ordering of different failure groups is introduced. These proposed methods are evaluated using a case study related to the edible bird nest industry in Sarawak, Malaysia. In short, the contributions of this paper are threefold: (1) a new Euclidean distance-based similarity measure, (2) a new risk interval measure for a group of failure modes, and (3) a new analysis of risk ordering of different failure groups. © 2014 The Natural Computing Applications Forum.

A noisy-smoothing relevance feedback method for content-based medical image retrieval

In this paper, we address a new problem of noisy images which present in the procedure of relevance feedback for medical image retrieval. We concentrate on the noisy images, caused by the users mislabeling some irrelevant images as relevant ones, and a noisy-smoothing relevance feedback (NS-RF) method is proposed. In NS-RF, a two-step strategy is proposed to handle the noisy images. In step 1, a noisy elimination algorithm is adopted to identify and eliminate the noisy images. In step 2, to further alleviate the influence of noisy images, a fuzzy membership function is employed to estimate the relevance probabilities of retained relevant images. After noisy handling, the fuzzy support vector machine, which can take into account different relevant images with different relevance probabilities, is adopted to re-rank the images. The experimental results on the IRMA medical image collection demonstrate that the proposed method can deal with the noisy images effectively.