Species-based and community-level approaches to conservation prioritization

One major reason for the global decline of biodiversity is habitat loss and fragmentation. Conservation areas can be designed to reduce biodiversity loss, but as resources are limited, conservation efforts need to be prioritized in order to achieve best possible outcomes. The field of systematic conservation planning developed as a response to opportunistic approaches to conservation that often resulted in biased representation of biological diversity. The last two decades have seen the development of increasingly sophisticated methods that account for information about biodiversity conservation goals (benefits), economical considerations (costs) and socio-political constraints. In this thesis I focus on two general topics related to systematic conservation planning. First, I address two aspects of the question about how biodiversity features should be valued. (i) I investigate the extremely important but often neglected issue of differential prioritization of species for conservation. Species prioritization can be based on various criteria, and is always goal-dependent, but can also be implemented in a scientifically more rigorous way than what is the usual practice. (ii) I introduce a novel framework for conservation prioritization, which is based on continuous benefit functions that convert increasing levels of biodiversity feature representation to increasing conservation value using the principle that more is better. Traditional target-based systematic conservation planning is a special case of this approach, in which a step function is used for the benefit function. We have further expanded the benefit function framework for area prioritization to address issues such as protected area size and habitat vulnerability. In the second part of the thesis I address the application of community level modelling strategies to conservation prioritization. One of the most serious issues in systematic conservation planning currently is not the deficiency of methodology for selection and design, but simply the lack of data. Community level modelling offers a surrogate strategy that makes conservation planning more feasible in data poor regions. We have reviewed the available community-level approaches to conservation planning. These range from simplistic classification techniques to sophisticated modelling and selection strategies. We have also developed a general and novel community level approach to conservation prioritization that significantly improves on methods that were available before. This thesis introduces further degrees of realism into conservation planning methodology. The benefit function -based conservation prioritization framework largely circumvents the problematic phase of target setting, and allowing for trade-offs between species representation provides a more flexible and hopefully more attractive approach to conservation practitioners. The community-level approach seems highly promising and should prove valuable for conservation planning especially in data poor regions. Future work should focus on integrating prioritization methods to deal with multiple aspects in combination influencing the prioritization process, and further testing and refining the community level strategies using real, large datasets.

Detecting actions of fruit flies

In this thesis we describe a system that tracks fruit flies in video and automatically detects and classifies their actions. We introduce Caltech Fly-vs-Fly Interactions, a new dataset that contains hours of video showing pairs of fruit flies engaging in social interactions, and is published with complete expert annotations and articulated pose trajectory features. We compare experimentally the value of a frame-level feature representation with the more elaborate notion of bout features that capture the structure within actions. Similarly, we compare a simple sliding window classifier architecture with a more sophisticated structured output architecture, and find that window based detectors outperform the much slower structured counterparts, and approach human performance. In addition we test the top performing detector on the CRIM13 mouse dataset, finding that it matches the performance of the best published method.

Augmented attribute representations

We propose a new learning method to infer a mid-level feature representation that combines the advantage of semantic attribute representations with the higher expressive power of non-semantic features. The idea lies in augmenting an existing attribute-based representation with additional dimensions for which an autoencoder model is coupled with a large-margin principle. This construction allows a smooth transition between the zero-shot regime with no training example, the unsupervised regime with training examples but without class labels, and the supervised regime with training examples and with class labels. The resulting optimization problem can be solved efficiently, because several of the necessity steps have closed-form solutions. Through extensive experiments we show that the augmented representation achieves better results in terms of object categorization accuracy than the semantic representation alone. © 2012 Springer-Verlag.

Gait Dynamics for Recognition and Classification

This paper describes a representation of the dynamics of human walking action for the purpose of person identification and classification by gait appearance. Our gait representation is based on simple features such as moments extracted from video silhouettes of human walking motion. We claim that our gait dynamics representation is rich enough for the task of recognition and classification. The use of our feature representation is demonstrated in the task of person recognition from video sequences of orthogonal views of people walking. We demonstrate the accuracy of recognition on gait video sequences collected over different days and times, and under varying lighting environments. In addition, preliminary results are shown on gender classification using our gait dynamics features.

Robust Multimodal Person Identification With Limited Training Data

This paper presents a novel method of audio-visual feature-level fusion for person identification where both the speech and facial modalities may be corrupted, and there is a lack of prior knowledge about the corruption. Furthermore, we assume there are limited amount of training data for each modality (e.g., a short training speech segment and a single training facial image for each person). A new multimodal feature representation and a modified cosine similarity are introduced to combine and compare bimodal features with limited training data, as well as vastly differing data rates and feature sizes. Optimal feature selection and multicondition training are used to reduce the mismatch between training and testing, thereby making the system robust to unknown bimodal corruption. Experiments have been carried out on a bimodal dataset created from the SPIDRE speaker recognition database and AR face recognition database with variable noise corruption of speech and occlusion in the face images. The system's speaker identification performance on the SPIDRE database, and facial identification performance on the AR database, is comparable with the literature. Combining both modalities using the new method of multimodal fusion leads to significantly improved accuracy over the unimodal systems, even when both modalities have been corrupted. The new method also shows improved identification accuracy compared with the bimodal systems based on multicondition model training or missing-feature decoding alone.

Machine learning for transient discovery in Pan-STARRS1 difference imaging

Efficient identification and follow-up of astronomical transients is hindered by the need for humans to manually select promising candidates from data streams that contain many false positives. These artefacts arise in the difference images that are produced by most major ground-based time-domain surveys with large format CCD cameras. This dependence on humans to reject bogus detections is unsustainable for next generation all-sky surveys and significant effort is now being invested to solve the problem computationally. In this paper, we explore a simple machine learning approach to real-bogus classification by constructing a training set from the image data of similar to 32 000 real astrophysical transients and bogus detections from the Pan-STARRS1 Medium Deep Survey. We derive our feature representation from the pixel intensity values of a 20 x 20 pixel stamp around the centre of the candidates. This differs from previous work in that it works directly on the pixels rather than catalogued domain knowledge for feature design or selection. Three machine learning algorithms are trained (artificial neural networks, support vector machines and random forests) and their performances are tested on a held-out subset of 25 per cent of the training data. We find the best results from the random forest classifier and demonstrate that by accepting a false positive rate of 1 per cent, the classifier initially suggests a missed detection rate of around 10 per cent. However, we also find that a combination of bright star variability, nuclear transients and uncertainty in human labelling means that our best estimate of the missed detection rate is approximately 6 per cent.

Analyse de mouvements faciaux à partir d'images vidéo

Lors d'une intervention conversationnelle, le langage est supporté par une communication non-verbale qui joue un rôle central dans le comportement social humain en permettant de la rétroaction et en gérant la synchronisation, appuyant ainsi le contenu et la signification du discours. En effet, 55% du message est véhiculé par les expressions faciales, alors que seulement 7% est dû au message linguistique et 38% au paralangage. L'information concernant l'état émotionnel d'une personne est généralement inférée par les attributs faciaux. Cependant, on ne dispose pas vraiment d'instruments de mesure spécifiquement dédiés à ce type de comportements. En vision par ordinateur, on s'intéresse davantage au développement de systèmes d'analyse automatique des expressions faciales prototypiques pour les applications d'interaction homme-machine, d'analyse de vidéos de réunions, de sécurité, et même pour des applications cliniques. Dans la présente recherche, pour appréhender de tels indicateurs observables, nous essayons d'implanter un système capable de construire une source consistante et relativement exhaustive d'informations visuelles, lequel sera capable de distinguer sur un visage les traits et leurs déformations, permettant ainsi de reconnaître la présence ou absence d'une action faciale particulière. Une réflexion sur les techniques recensées nous a amené à explorer deux différentes approches. La première concerne l'aspect apparence dans lequel on se sert de l'orientation des gradients pour dégager une représentation dense des attributs faciaux. Hormis la représentation faciale, la principale difficulté d'un système, qui se veut être général, est la mise en œuvre d'un modèle générique indépendamment de l'identité de la personne, de la géométrie et de la taille des visages. La démarche qu'on propose repose sur l'élaboration d'un référentiel prototypique à partir d'un recalage par SIFT-flow dont on démontre, dans cette thèse, la supériorité par rapport à un alignement conventionnel utilisant la position des yeux. Dans une deuxième approche, on fait appel à un modèle géométrique à travers lequel les primitives faciales sont représentées par un filtrage de Gabor. Motivé par le fait que les expressions faciales sont non seulement ambigües et incohérentes d'une personne à une autre mais aussi dépendantes du contexte lui-même, à travers cette approche, on présente un système personnalisé de reconnaissance d'expressions faciales, dont la performance globale dépend directement de la performance du suivi d'un ensemble de points caractéristiques du visage. Ce suivi est effectué par une forme modifiée d'une technique d'estimation de disparité faisant intervenir la phase de Gabor. Dans cette thèse, on propose une redéfinition de la mesure de confiance et introduisons une procédure itérative et conditionnelle d'estimation du déplacement qui offrent un suivi plus robuste que les méthodes originales.

Learning enhanced 3D models for vehicle tracking

This paper presents an enhanced hypothesis verification strategy for 3D object recognition. A new learning methodology is presented which integrates the traditional dichotomic object-centred and appearance-based representations in computer vision giving improved hypothesis verification under iconic matching. The "appearance" of a 3D object is learnt using an eigenspace representation obtained as it is tracked through a scene. The feature representation implicitly models the background and the objects observed enabling the segmentation of the objects from the background. The method is shown to enhance model-based tracking, particularly in the presence of clutter and occlusion, and to provide a basis for identification. The unified approach is discussed in the context of the traffic surveillance domain. The approach is demonstrated on real-world image sequences and compared to previous (edge-based) iconic evaluation techniques.

Rede neural não supervisionada baseada em transformada Karhunen-Loève para processamento, compressão e reconhecimento de imagens

The merit of the Karhunen-Loève transform is well known. Since its basis is the eigenvector set of the covariance matrix, a statistical, not functional, representation of the variance in pattern ensembles is generated. By using the Karhunen-Loève transform coefficients as a natural feature representation of a character image, the eigenvector set can be regarded as an feature extractor for a classifier.

Modeling Features at Runtime

A feature represents a functional requirement fulfilled by a system. Since many maintenance tasks are expressed in terms of features, it is important to establish the correspondence between a feature and its implementation in source code. Traditional approaches to establish this correspondence exercise features to generate a trace of runtime events, which is then processed by post-mortem analysis. These approaches typically generate large amounts of data to analyze. Due to their static nature, these approaches do not support incremental and interactive analysis of features. We propose a radically different approach called live feature analysis, which provides a model at runtime of features. Our approach analyzes features on a running system and also makes it possible to grow feature representations by exercising different scenarios of the same feature, and identifies execution elements even to the sub-method level. We describe how live feature analysis is implemented effectively by annotating structural representations of code based on abstract syntax trees. We illustrate our live analysis with a case study where we achieve a more complete feature representation by exercising and merging variants of feature behavior and demonstrate the efficiency or our technique with benchmarks.

Improving classification using a Confidence Matrix based on weak classifiers applied to OCR

This paper proposes a new feature representation method based on the construction of a Confidence Matrix (CM). This representation consists of posterior probability values provided by several weak classifiers, each one trained and used in different sets of features from the original sample. The CM allows the final classifier to abstract itself from discovering underlying groups of features. In this work the CM is applied to isolated character image recognition, for which several set of features can be extracted from each sample. Experimentation has shown that the use of CM permits a significant improvement in accuracy in most cases, while the others remain the same. The results were obtained after experimenting with four well-known corpora, using evolved meta-classifiers with the k-Nearest Neighbor rule as a weak classifier and by applying statistical significance tests.

Marine mammal species detection and classification

Thesis (Ph.D.)--University of Washington, 2016-04

Automatically extracting polarity-bearing topics for cross-domain sentiment classification

Joint sentiment-topic (JST) model was previously proposed to detect sentiment and topic simultaneously from text. The only supervision required by JST model learning is domain-independent polarity word priors. In this paper, we modify the JST model by incorporating word polarity priors through modifying the topic-word Dirichlet priors. We study the polarity-bearing topics extracted by JST and show that by augmenting the original feature space with polarity-bearing topics, the in-domain supervised classifiers learned from augmented feature representation achieve the state-of-the-art performance of 95% on the movie review data and an average of 90% on the multi-domain sentiment dataset. Furthermore, using feature augmentation and selection according to the information gain criteria for cross-domain sentiment classification, our proposed approach performs either better or comparably compared to previous approaches. Nevertheless, our approach is much simpler and does not require difficult parameter tuning.

Connecting social media to e-commerce:cold-start product recommendation using microblogging information

In recent years, the boundaries between e-commerce and social networking have become increasingly blurred. Many e-commerce websites support the mechanism of social login where users can sign on the websites using their social network identities such as their Facebook or Twitter accounts. Users can also post their newly purchased products on microblogs with links to the e-commerce product web pages. In this paper, we propose a novel solution for cross-site cold-start product recommendation, which aims to recommend products from e-commerce websites to users at social networking sites in 'cold-start' situations, a problem which has rarely been explored before. A major challenge is how to leverage knowledge extracted from social networking sites for cross-site cold-start product recommendation. We propose to use the linked users across social networking sites and e-commerce websites (users who have social networking accounts and have made purchases on e-commerce websites) as a bridge to map users' social networking features to another feature representation for product recommendation. In specific, we propose learning both users' and products' feature representations (called user embeddings and product embeddings, respectively) from data collected from e-commerce websites using recurrent neural networks and then apply a modified gradient boosting trees method to transform users' social networking features into user embeddings. We then develop a feature-based matrix factorization approach which can leverage the learnt user embeddings for cold-start product recommendation. Experimental results on a large dataset constructed from the largest Chinese microblogging service Sina Weibo and the largest Chinese B2C e-commerce website JingDong have shown the effectiveness of our proposed framework.

Hierarchically grouped 2D local features applied to edge contour localisation

One of the most significant research topics in computer vision is object detection. Most of the reported object detection results localise the detected object within a bounding box, but do not explicitly label the edge contours of the object. Since object contours provide a fundamental diagnostic of object shape, some researchers have initiated work on linear contour feature representations for object detection and localisation. However, linear contour feature-based localisation is highly dependent on the performance of linear contour detection within natural images, and this can be perturbed significantly by a cluttered background. In addition, the conventional approach to achieving rotation-invariant features is to rotate the feature receptive field to align with the local dominant orientation before computing the feature representation. Grid resampling after rotation adds extra computational cost and increases the total time consumption for computing the feature descriptor. Though it is not an expensive process if using current computers, it is appreciated that if each step of the implementation is faster to compute especially when the number of local features is increasing and the application is implemented on resource limited ”smart devices”, such as mobile phones, in real-time. Motivated by the above issues, a 2D object localisation system is proposed in this thesis that matches features of edge contour points, which is an alternative method that takes advantage of the shape information for object localisation. This is inspired by edge contour points comprising the basic components of shape contours. In addition, edge point detection is usually simpler to achieve than linear edge contour detection. Therefore, the proposed localization system could avoid the need for linear contour detection and reduce the pathological disruption from the image background. Moreover, since natural images usually comprise many more edge contour points than interest points (i.e. corner points), we also propose new methods to generate rotation-invariant local feature descriptors without pre-rotating the feature receptive field to improve the computational efficiency of the whole system. In detail, the 2D object localisation system is achieved by matching edge contour points features in a constrained search area based on the initial pose-estimate produced by a prior object detection process. The local feature descriptor obtains rotation invariance by making use of rotational symmetry of the hexagonal structure. Therefore, a set of local feature descriptors is proposed based on the hierarchically hexagonal grouping structure. Ultimately, the 2D object localisation system achieves a very promising performance based on matching the proposed features of edge contour points with the mean correct labelling rate of the edge contour points 0.8654 and the mean false labelling rate 0.0314 applied on the data from Amsterdam Library of Object Images (ALOI). Furthermore, the proposed descriptors are evaluated by comparing to the state-of-the-art descriptors and achieve competitive performances in terms of pose estimate with around half-pixel pose error.