Irene Cheng

Bio: Irene Cheng is an academic researcher from University of Alberta. The author has contributed to research in topics: Visualization & Image segmentation. The author has an hindex of 23, co-authored 199 publications receiving 2154 citations. Previous affiliations of Irene Cheng include École Centrale Paris & University of Pennsylvania.

Generalized Random Walks for Fusion of Multi-Exposure Images

Abstract: A single captured image of a real-world scene is usually insufficient to reveal all the details due to under- or over-exposed regions To solve this problem, images of the same scene can be first captured under different exposure settings and then combined into a single image using image fusion techniques In this paper, we propose a novel probabilistic model-based fusion technique for multi-exposure images Unlike previous multi-exposure fusion methods, our method aims to achieve an optimal balance between two quality measures, ie, local contrast and color consistency, while combining the scene details revealed under different exposures A generalized random walks framework is proposed to calculate a globally optimal solution subject to the two quality measures by formulating the fusion problem as probability estimation Experiments demonstrate that our algorithm generates high-quality images at low computational cost Comparisons with a number of other techniques show that our method generates better results in most cases

Fluid Vector Flow and Applications in Brain Tumor Segmentation

Abstract: In this paper, we propose a new approach that we call the ldquofluid vector flowrdquo (FVF) active contour model to address problems of insufficient capture range and poor convergence for concavities. With the ability to capture a large range and extract concave shapes, FVF demonstrates improvements over techniques like gradient vector flow, boundary vector flow, and magnetostatic active contour on three sets of experiments: synthetic images, pediatric head MRI images, and brain tumor MRI images from the Internet brain segmentation repository.

Cross-Scale Coefficient Selection for Volumetric Medical Image Fusion

Abstract: Joint analysis of medical data collected from different imaging modalities has become a common clinical practice. Therefore, image fusion techniques, which provide an efficient way of combining and enhancing information, have drawn increasing attention from the medical community. In this paper, we propose a novel cross-scale fusion rule for multiscale-decomposition-based fusion of volumetric medical images taking into account both intrascale and interscale consistencies. An optimal set of coefficients from the multiscale representations of the source images is determined by effective exploitation of neighborhood information. An efficient color fusion scheme is also proposed. Experiments demonstrate that our fusion rule generates better results than existing rules.

Quality metric for approximating subjective evaluation of 3-D objects

Abstract: Many factors, such as the number of vertices and the resolution of texture, can affect the display quality of three-dimensional (3-D) objects. When the resources of a graphics system are not sufficient to render the ideal image, degradation is inevitable. It is, therefore, important to study how individual factors will affect the overall quality, and how the degradation can be controlled given limited resources. In this paper, the essential factors determining the display quality are reviewed. We then integrate two important ones, resolution of texture and resolution of wireframe, and use them in our model as a perceptual metric. We assess this metric using statistical data collected from a 3-D quality evaluation experiment. The statistical model and the methodology to assess the display quality metric are discussed. A preliminary study of the reliability of the estimates is also described. The contribution of this paper lies in: 1) determining the relative importance of wireframe versus texture resolution in perceptual quality evaluation and 2) proposing an experimental strategy for verifying and fitting a quantitative model that estimates 3-D perceptual quality. The proposed quantitative method is found to fit closely to subjective ratings by human observers based on preliminary experimental results.

Method and apparatus for high resolution 3D scanning

Abstract: A method and apparatus for fast high resolution 3D scanning of objects possibly with holes in them includes providing an imaging device, at least one laser pattern projector, sensors adapted to sense a position on an object of a laser pattern projected by the laser pattern projector, sensors adapted to sense the exact identity of the laser patterns that did not fall on the object being scanned, and multiple independent imaging systems coupled with light interference eliminators designed for simultaneously scanning depth and texture data on a 3D object. A computer processor is provided which is adapted to receive from the imaging device a scanned image of an object and adapted to receive from the sensors data regarding the position on the object of the laser pattern projected by the laser pattern projector. The computer processor integrates and registers data from one or more independent imaging systems and sensors to create a high resolution 3D image with accurate depth and texture details.

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Standards for educational and psychological testing

Abstract: Pedagosko i psiholosko testiranje i procjenjivanje spadaju među najvažnije doprinose znanosti o ponasanju nasem drustvu i pružaju temeljna i znacajna poboljsanja u odnosu na ranije postupke. Iako se ne može ustvrditi da su svi testovi dovoljno usavrseni niti da su sva testiranja razborita i korisna, postoji velika kolicina informacija koje ukazuju na ucinkovitost kvalitetnih instrumenata u onim situacijama u kojima je njihovo koristenje potkrijepljeno validacijskim podacima. Pravilna upotreba testova može dovesti do boljih odluka o pojedincima i programima nego sto bi to bio slucaj bez njihovog koristenja, a također i ukazati na put za siri i pravedniji pristup obrazovanju i zaposljavanju. Međutim, losa upotreba testova može dovesti do zamjetne stete nanesene ispitanicima i drugim sudionicima u procesu donosenja odluka na temelju testovnih podataka. Cilj Standarda je promoviranje kvalitetne i eticne upotrebe testova te uspostavljanje osnovice za ocjenu kvalitete postupaka testiranja. Svrha objavljivanja Standarda je uspostavljanje kriterija za evaluaciju testova, provedbe testiranja i posljedica upotrebe testova. Iako bi evaluacija prikladnosti testa ili njegove primjene trebala ovisiti prvenstveno o strucnim misljenjima, Standardi pružaju okvir koji osigurava obuhvacanje svih relevantnih pitanja. Bilo bi poželjno da svi autori, sponzori, nakladnici i korisnici profesionalnih testova usvoje Standarde te da poticu druge da ih također prihvate.

Image Processing

Abstract: MUCKE aims to mine a large volume of images, to structure them conceptually and to use this conceptual structuring in order to improve large-scale image retrieval. The last decade witnessed important progress concerning low-level image representations. However, there are a number problems which need to be solved in order to unleash the full potential of image mining in applications. The central problem with low-level representations is the mismatch between them and the human interpretation of image content. This problem can be instantiated, for instance, by the incapability of existing descriptors to capture spatial relationships between the concepts represented or by their incapability to convey an explanation of why two images are similar in a content-based image retrieval framework. We start by assessing existing local descriptors for image classification and by proposing to use co-occurrence matrices to better capture spatial relationships in images. The main focus in MUCKE is on cleaning large scale Web image corpora and on proposing image representations which are closer to the human interpretation of images. Consequently, we introduce methods which tackle these two problems and compare results to state of the art methods. Note: some aspects of this deliverable are withheld at this time as they are pending review. Please contact the authors for a preview.