Lighting invariant urban street classification

In this paper we propose the hybrid use of illuminant invariant and RGB images to perform image classification of urban scenes despite challenging variation in lighting conditions. Coping with lighting change (and the shadows thereby invoked) is a non-negotiable requirement for long term autonomy using vision. One aspect of this is the ability to reliably classify scene components in the presence of marked and often sudden changes in lighting. This is the focus of this paper. Posed with the task of classifying all parts in a scene from a full colour image, we propose that lighting invariant transforms can reduce the variability of the scene, resulting in a more reliable classification. We leverage the ideas of “data transfer” for classification, beginning with full colour images for obtaining candidate scene-level matches using global image descriptors. This is commonly followed by superpixellevel matching with local features. However, we show that if the RGB images are subjected to an illuminant invariant transform before computing the superpixel-level features, classification is significantly more robust to scene illumination effects. The approach is evaluated using three datasets. The first being our own dataset and the second being the KITTI dataset using manually generated ground truth for quantitative analysis. We qualitatively evaluate the method on a third custom dataset over a 750m trajectory.

Machine learning for activity recognition : hip versus wrist data

Problem addressed Wrist-worn accelerometers are associated with greater compliance. However, validated algorithms for predicting activity type from wrist-worn accelerometer data are lacking. This study compared the activity recognition rates of an activity classifier trained on acceleration signal collected on the wrist and hip. Methodology 52 children and adolescents (mean age 13.7 +/- 3.1 year) completed 12 activity trials that were categorized into 7 activity classes: lying down, sitting, standing, walking, running, basketball, and dancing. During each trial, participants wore an ActiGraph GT3X+ tri-axial accelerometer on the right hip and the non-dominant wrist. Features were extracted from 10-s windows and inputted into a regularized logistic regression model using R (Glmnet + L1). Results Classification accuracy for the hip and wrist was 91.0% +/- 3.1% and 88.4% +/- 3.0%, respectively. The hip model exhibited excellent classification accuracy for sitting (91.3%), standing (95.8%), walking (95.8%), and running (96.8%); acceptable classification accuracy for lying down (88.3%) and basketball (81.9%); and modest accuracy for dance (64.1%). The wrist model exhibited excellent classification accuracy for sitting (93.0%), standing (91.7%), and walking (95.8%); acceptable classification accuracy for basketball (86.0%); and modest accuracy for running (78.8%), lying down (74.6%) and dance (69.4%). Potential Impact Both the hip and wrist algorithms achieved acceptable classification accuracy, allowing researchers to use either placement for activity recognition.

Pollen image classification using the classifynder system: Algorithm comparison and a case study on New Zealand honey

We describe an investigation into how Massey University’s Pollen Classifynder can accelerate the understanding of pollen and its role in nature. The Classifynder is an imaging microscopy system that can locate, image and classify slide based pollen samples. Given the laboriousness of purely manual image acquisition and identification it is vital to exploit assistive technologies like the Classifynder to enable acquisition and analysis of pollen samples. It is also vital that we understand the strengths and limitations of automated systems so that they can be used (and improved) to compliment the strengths and weaknesses of human analysts to the greatest extent possible. This article reviews some of our experiences with the Classifynder system and our exploration of alternative classifier models to enhance both accuracy and interpretability. Our experiments in the pollen analysis problem domain have been based on samples from the Australian National University’s pollen reference collection (2,890 grains, 15 species) and images bundled with the Classifynder system (400 grains, 4 species). These samples have been represented using the Classifynder image feature set.We additionally work through a real world case study where we assess the ability of the system to determine the pollen make-up of samples of New Zealand honey. In addition to the Classifynder’s native neural network classifier, we have evaluated linear discriminant, support vector machine, decision tree and random forest classifiers on these data with encouraging results. Our hope is that our findings will help enhance the performance of future releases of the Classifynder and other systems for accelerating the acquisition and analysis of pollen samples.

A comparison of classification algorithms within the Classifynder pollen imaging system

We describe an investigation into how Massey University's Pollen Classifynder can accelerate the understanding of pollen and its role in nature. The Classifynder is an imaging microscopy system that can locate, image and classify slide based pollen samples. Given the laboriousness of purely manual image acquisition and identification it is vital to exploit assistive technologies like the Classifynder to enable acquisition and analysis of pollen samples. It is also vital that we understand the strengths and limitations of automated systems so that they can be used (and improved) to compliment the strengths and weaknesses of human analysts to the greatest extent possible. This article reviews some of our experiences with the Classifynder system and our exploration of alternative classifier models to enhance both accuracy and interpretability. Our experiments in the pollen analysis problem domain have been based on samples from the Australian National University's pollen reference collection (2890 grains, 15 species) and images bundled with the Classifynder system (400 grains, 4 species). These samples have been represented using the Classifynder image feature set. In addition to the Classifynder's native neural network classifier, we have evaluated linear discriminant, support vector machine, decision tree and random forest classifiers on these data with encouraging results. Our hope is that our findings will help enhance the performance of future releases of the Classifynder and other systems for accelerating the acquisition and analysis of pollen samples. © 2013 AIP Publishing LLC.

Injury narrative text classification using factorization model

Narrative text is a useful way of identifying injury circumstances from the routine emergency department data collections. Automatically classifying narratives based on machine learning techniques is a promising technique, which can consequently reduce the tedious manual classification process. Existing works focus on using Naive Bayes which does not always offer the best performance. This paper proposes the Matrix Factorization approaches along with a learning enhancement process for this task. The results are compared with the performance of various other classification approaches. The impact on the classification results from the parameters setting during the classification of a medical text dataset is discussed. With the selection of right dimension k, Non Negative Matrix Factorization-model method achieves 10 CV accuracy of 0.93.

Detection of rain in acoustic recordings of the environment using machine learning techniques

This thesis is concerned with the detection and prediction of rain in environmental recordings using different machine learning algorithms. The results obtained in this research will help ecologists to efficiently analyse environmental data and monitor biodiversity.

Classification of activities of daily living of individual with limb loss

The monitoring of the actual activities of daily living of individuals with lower limb amputation is essential for an evidence-based fitting of the prosthesis, more particularly the choice of components (e.g., knees, ankles, feet)[1-4]. The purpose of this presentation was to give an overview of the categorization of the load regime data to assess the functional output and usage of the prosthesis of lower limb amputees has presented in several publications[5, 6]. The objectives were to present a categorization of load regime and to report the results for a case.

Centroid training to achieve effective text classification

Traditional text classification technology based on machine learning and data mining techniques has made a big progress. However, it is still a big problem on how to draw an exact decision boundary between relevant and irrelevant objects in binary classification due to much uncertainty produced in the process of the traditional algorithms. The proposed model CTTC (Centroid Training for Text Classification) aims to build an uncertainty boundary to absorb as many indeterminate objects as possible so as to elevate the certainty of the relevant and irrelevant groups through the centroid clustering and training process. The clustering starts from the two training subsets labelled as relevant or irrelevant respectively to create two principal centroid vectors by which all the training samples are further separated into three groups: POS, NEG and BND, with all the indeterminate objects absorbed into the uncertain decision boundary BND. Two pairs of centroid vectors are proposed to be trained and optimized through the subsequent iterative multi-learning process, all of which are proposed to collaboratively help predict the polarities of the incoming objects thereafter. For the assessment of the proposed model, F1 and Accuracy have been chosen as the key evaluation measures. We stress the F1 measure because it can display the overall performance improvement of the final classifier better than Accuracy. A large number of experiments have been completed using the proposed model on the Reuters Corpus Volume 1 (RCV1) which is important standard dataset in the field. The experiment results show that the proposed model has significantly improved the binary text classification performance in both F1 and Accuracy compared with three other influential baseline models.

Towards robust automatic affective classification of images using facial expressions for practical applications

Affect is an important feature of multimedia content and conveys valuable information for multimedia indexing and retrieval. Most existing studies for affective content analysis are limited to low-level features or mid-level representations, and are generally criticized for their incapacity to address the gap between low-level features and high-level human affective perception. The facial expressions of subjects in images carry important semantic information that can substantially influence human affective perception, but have been seldom investigated for affective classification of facial images towards practical applications. This paper presents an automatic image emotion detector (IED) for affective classification of practical (or non-laboratory) data using facial expressions, where a lot of “real-world” challenges are present, including pose, illumination, and size variations etc. The proposed method is novel, with its framework designed specifically to overcome these challenges using multi-view versions of face and fiducial point detectors, and a combination of point-based texture and geometry. Performance comparisons of several key parameters of relevant algorithms are conducted to explore the optimum parameters for high accuracy and fast computation speed. A comprehensive set of experiments with existing and new datasets, shows that the method is effective despite pose variations, fast, and appropriate for large-scale data, and as accurate as the method with state-of-the-art performance on laboratory-based data. The proposed method was also applied to affective classification of images from the British Broadcast Corporation (BBC) in a task typical for a practical application providing some valuable insights.

Petrology and textural classification of the Jericho kimberlite, northern Slave Province, Nunavut, Canada

The paper presents data on petrology, bulk rock and mineral compositions, and textural classification of the Middle Jurassic Jericho kimberlite (Slave craton, Canada). The kimberlite was emplaced as three steep-sided pipes in granite that was overlain by limestones and minor soft sediments. The pipes are infilled with hypabyssal and pyroclastic kimberlites and connected to a satellite pipe by a dyke. The Jericho kimberlite is classified as a Group Ia, lacking groundmass tetraferriphlogopite and containing monticellite pseudomorphs. The kimberlite formed, during several consecutive emplacement events of compositionally different batches of kimberlite magma. Core-logging and thin-section observations identified at least two phases of hypabyssal kimberlites and three phases of pyroclastic kimberlites. Hypabyssal kimberlites intruded as a main dyke (HK1) and as late small-volume aphanitic and vesicular dykes. Massive pyroclastic kimberlite (MPK1) predominantly filled the northern and southern lobes of the pipe and formed from magma different from the HK1 magma. The MPK1 magma crystallized Ti-, Fe-, and Cr-rich phlogopite without rims of barian phlogopite, and clinopyroxene and spinel without atoll structures. MPK1 textures, superficially reminiscent of tuffisitic kimberlite, are caused by pervasive contamination by granite xenoliths. The next explosive events filled the central lobe with two varieties of pyroclastic kimberlite: (1) massive and (2) weakly bedded, normally graded pyroclastic kimberlite. The geology of the Jericho pipe differs from the geology of South African or the Prairie kimberlites, but may resemble Lac de Gras pipes, in which deeper erosion removed upper fades of resedimented kimberlites.

Using decision trees to understand structure in missing data

Objectives Demonstrate the application of decision trees – classification and regression trees (CARTs), and their cousins, boosted regression trees (BRTs) – to understand structure in missing data. Setting Data taken from employees at three different industry sites in Australia. Participants 7915 observations were included. Materials and Methods The approach was evaluated using an occupational health dataset comprising results of questionnaires, medical tests, and environmental monitoring. Statistical methods included standard statistical tests and the ‘rpart’ and ‘gbm’ packages for CART and BRT analyses, respectively, from the statistical software ‘R’. A simulation study was conducted to explore the capability of decision tree models in describing data with missingness artificially introduced. Results CART and BRT models were effective in highlighting a missingness structure in the data, related to the Type of data (medical or environmental), the site in which it was collected, the number of visits and the presence of extreme values. The simulation study revealed that CART models were able to identify variables and values responsible for inducing missingness. There was greater variation in variable importance for unstructured compared to structured missingness. Discussion Both CART and BRT models were effective in describing structural missingness in data. CART models may be preferred over BRT models for exploratory analysis of missing data, and selecting variables important for predicting missingness. BRT models can show how values of other variables influence missingness, which may prove useful for researchers. Conclusion Researchers are encouraged to use CART and BRT models to explore and understand missing data.

Multiple stages classification of Alzheimer’s disease based on structural brain networks using generalized low rank approximations (GLRAM)

To classify each stage for a progressing disease such as Alzheimer’s disease is a key issue for the disease prevention and treatment. In this study, we derived structural brain networks from diffusion-weighted MRI using whole-brain tractography since there is growing interest in relating connectivity measures to clinical, cognitive, and genetic data. Relatively little work has usedmachine learning to make inferences about variations in brain networks in the progression of the Alzheimer’s disease. Here we developed a framework to utilize generalized low rank approximations of matrices (GLRAM) and modified linear discrimination analysis for unsupervised feature learning and classification of connectivity matrices. We apply the methods to brain networks derived from DWI scans of 41 people with Alzheimer’s disease, 73 people with EMCI, 38 people with LMCI, 47 elderly healthy controls and 221 young healthy controls. Our results show that this new framework can significantly improve classification accuracy when combining multiple datasets; this suggests the value of using data beyond the classification task at hand to model variations in brain connectivity.

Cross database training of audio-visual hidden Markov models for phone recognition

Speech recognition can be improved by using visual information in the form of lip movements of the speaker in addition to audio information. To date, state-of-the-art techniques for audio-visual speech recognition continue to use audio and visual data of the same database for training their models. In this paper, we present a new approach to make use of one modality of an external dataset in addition to a given audio-visual dataset. By so doing, it is possible to create more powerful models from other extensive audio-only databases and adapt them on our comparatively smaller multi-stream databases. Results show that the presented approach outperforms the widely adopted synchronous hidden Markov models (HMM) trained jointly on audio and visual data of a given audio-visual database for phone recognition by 29% relative. It also outperforms the external audio models trained on extensive external audio datasets and also internal audio models by 5.5% and 46% relative respectively. We also show that the proposed approach is beneficial in noisy environments where the audio source is affected by the environmental noise.

Combined NIR/MIR analysis: A novel method for the classification of complex substances such as Illicium verum Hook. F. and its adulterants

A novel combined near- and mid-infrared (NIR and MIR) spectroscopic method has been researched and developed for the analysis of complex substances such as the Traditional Chinese Medicine (TCM), Illicium verum Hook. F. (IVHF), and its noxious adulterant, Iuicium lanceolatum A.C. Smith (ILACS). Three types of spectral matrix were submitted for classification with the use of the linear discriminant analysis (LDA) method. The data were pretreated with either the successive projections algorithm (SPA) or the discrete wavelet transform (DWT) method. The SPA method performed somewhat better, principally because it required less spectral features for its pretreatment model. Thus, NIR or MIR matrix as well as the combined NIR/MIR one, were pretreated by the SPA method, and then analysed by LDA. This approach enabled the prediction and classification of the IVHF, ILACS and mixed samples. The MIR spectral data produced somewhat better classification rates than the NIR data. However, the best results were obtained from the combined NIR/MIR data matrix with 95–100% correct classifications for calibration, validation and prediction. Principal component analysis (PCA) of the three types of spectral data supported the results obtained with the LDA classification method.