In this paper, we describe the model for faculty diversity developed as part of the Professorial Advancement Initiative (PAI) funded under the NSF AGEP program. The PAI, consisting of 12 of the 14 Big Ten Academic Alliance universities,1 had the goal of doubling the rate at which the universities hired tenure-track minoritized faculty, defined by National Science Foundation as African Americans, Hispanic/Latinx, Native Americans, and Pacific Islanders. This paper reviews the key programmatic elements of the PAI and discusses lessons learned and the practices developed that helped the Alliance achieve its faculty diversity goal.

The postdoctoral pathway to the professoriate is an important source of future faculty talent. This paper focuses on underrepresented minority (URM) postdocs and the challenges they face as they prepare for tenure track positions in the academy. To date, much of the research on URM success in science, technology, engineering, and mathematics (STEM) fields has focused on student interest in STEM fields and STEM curricula, both at the K-12 and undergraduate levels. By comparison, very little attention has been given to studying the career issues of URM STEM postdocs seeking to enter the professoriate, especially from a critical race lens. The postdocs who participated in our study completed an interview, which probed self-efficacy, sense of belonging, social identity, and challenges in the field. After coding and the analysis, 26 sub-themes emerged including feelings about the culture of their work environment and perceptions of stereotypes and biases. The results of the study, which are broadly applicable, have improved our understanding of the factors affecting the progression of URM postdocs into the professoriate and have informed mentor training and career success strategies for URM postdocs.

This paper surveys various Image denoising parameters by Fractal Image denoising [1], Neighbouring wavelet Coefficients [2] & Directional Filter Banks [3]. The survey analysis the RMSE,PSNR values obtained with various Image denoising techniques including Predictive FW Scheme, Predictive quadtree-based FW scheme with Collage error decomposition criterion, Predictive pixel based fractal scheme with uniform partitioning , Predictive pixel domain

In practice, training language models for individual authors is often expensive because of limited data resources. In such cases, Neural Network Language Models (NNLMs), generally outperform the traditional non-parametric N-gram models. Here we investigate the performance of a feed-forward NNLM on an authorship attribution problem, with moderate author set size and relatively limited data. We also consider how the text topics impact performance. Compared with a well-constructed N-gram baseline method with Kneser-Ney smoothing, the proposed method achieves nearly 2.5% reduction in perplexity and increases author classification accuracy by 3.43% on average, given as few as 5 test sentences. The performance is very competitive with the state of the art in terms of accuracy and demand on test data. The source code, preprocessed datasets, a detailed description of the methodology and results are available at https: //github.com/zge/authorship-attribution. Introduction Authorship attribution refers to identifying authors from given texts by their unique textual features. It is challenging since the author’s style may vary from time to time by topics, mood and environment. Many methods have been explored to address this problem, such as Latent Dirichlet Allocation for topic modeling (Seroussi, Zukerman, and Bohnert 2011) and Naive Bayes for text classification (Coyotl-Morales et al. 2006). Regarding language modeling methods, there is mixed advocacy for the conventional N-gram methods (Kešelj et al. 2003) and methods using more compact and distributed representations, like Neural Network Language Models (NNLMs), which was claimed to capture semantics better with limited training data (Bengio et al. 2003). Most NNLM toolkits available (Mikolov et al. 2010) are designed for recurrent NNLMs which are better for capturing complex and longer text patterns and require more training data. In contrast, the feed-forward NNLM framework we proposed is less computationally expensive and more suitable for language modeling with limited data. It is developed in MATLAB with full network tuning functionalities. The database we use is composed of transcripts of 16 video courses taken from Coursera, collected one sentence per line into a text file for each course. To reduce the influence of “topic” on author/instructor classification, courses were all selected from science and engineering fields, such Copyright c © 2015, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. as Algorithm, DSP, Data Mining, IT, Machine Learning, NLP, etc. There are 8000+ sentences/course and about 20 words/sentence on average. The vocabulary size of each author varies from 3000 to 9000. After stemming with Porter’s algorithm and pruning words with frequency less than 1/100, 000, author vocabulary size is reduced to a range from 1800 to 2700, with average size around 2000. Fig. 1 shows the vocabulary size for each course, under various conditions and the database coverage with the most frequent k words (k = 500, 1000, 2000) after stemming and pruning. 0 2 4 6 8 10 12 14 16 0 500

Aluminum Oxynitride (ALON (R) Optical Ceramic) combines broadband transparency with excellent mechanical properties. ALON's cubic structure means that it is transparent in its polycrystalline form, allowing it to be manufactured by conventional powder processing techniques. Surmet controls every aspect of the manufacturing process, beginning with synthesis of ALON (R) powder, continuing through forming/heat treatment of blanks, ending with optical fabrication of ALON (R) windows. Surmet has made significant progress in its production capability in recent years. Additional scale up of Surmet's manufacturing capability, for complex geometries, larger sizes and higher quantities, is underway.The requirements for modern aircraft are driving the need for conformal windows for future sensor systems. However, limitations on optical systems and the ability to produce windows in complex geometries currently limit the geometry of existing windows and window assemblies to faceted assemblies of flat windows. Surmet's ability to produce large curved ALON (R) blanks is an important step in the development of conformal windows for future aircraft applications.

In conventional subband/wavelet image coding, the subband decomposition is performed on the spatial-domain image. Here, we introduce a novel decomposition where the subband decomposition is performed on the global DCT spectrum of the image. That is, the two-dimensional spectrum rather than the image is represented by a sum of basis functions, each weighted by the transform coefficients. The distinct features of this decomposition are analyzed from a transform perspective. This spectral subband decomposition is then used as the basis for a new image coder, building on the condensed wavelet packet (CWP) algorithm. Ironically, this new method is shown to have lower arithmetic complexity than conventional subband/wavelet coders that directly decompose a time or spatial domain signal. Comparisons of the new method against conventional subband/wavelet coders that use the popular 9/7 dyadic decomposition, condensed wavelet packets, and generalized lapped orthogonal transforms, show that the new method has lower complexity and higher compression performance.

The map is more important than ever, but what it can represent and how it is delivered are changing radically as new geospatial technologies emerge. In the field of geology, paper maps have always aimed to represent the complex three-dimensional world beneath our feet and make it understandable to us at the surface. However, new smart phone mapping applications enable us to take the map with us more easily and to ask questions of it wherever we are, and to add our own observations on to it. Digital survey and modelling technologies will enable geologists to communicate geology in 3D for us, rather than having to translate it to 2D; everyone will then see the geology as the geologist does as we take visualisation from the lab to the street. As an interactive tool the geological map of the future will be very different.

This article analyses the changing position of gender in the European Employment Strategy (EES) since its 2005 relaunch. Overall, we find a picture of mixed progress towards gender equality goals across Member States. There is evidence of the EU soft law approach leading to positive developments as the use of targets in conjunction with Country-Specific Recommendations and Points-to-Watch have had some influence in promoting gender equality policies among Member States. However, the weakened position of gender mainstreaming in European-level initiatives has led to gender being marginalised or ignored in national and EU policy responses to the crisis. The prominence of gender has declined further in the 2010 revision of the EES under the 2020 banner. This introduces new risks as the emphasis on gender equality falls further down the list of priorities in the streamlining of the Lisbon Process.

A novel approach pertaining to the fast and efficient retrieval and storage of video sequences utilizing MPEG-1/2 motion vectors is presented in this paper. A clip first must be segmented into consistent video segments based on the basic editing effects - cuts, dissolves, and wipes. All group of pictures (GOPs) are then extracted from the clip and decomposed further into I-frames and P-frames (B-frames are disregarded). The initial frame of the sequence is manually segmented into objects, and the selected objects are automatically tracked through the entire sequence using MPEG-1/2 motion vectors. Features pertaining to the edge histogram are extracted from each tracked object (I/P frames only) and then maintained with the associated frame. An example image or video clip is then presented to the system and the 5 best matching images are retrieved. Results are shown for the standard video sequences such as foreman, tennis, etc.

Segmenting videos into meaningful real-world objects remains one of the most challenging image processing research topics. Frame-by-frame object tracking is especially challenging due to the limitations of existing image segmentation algorithms often resulting in inconsistent regions occurring between adjacent frames. This work addresses this problem by introducing an innovative region matching and validation algorithm of segmented objects extracted from adjacent frames. Motion vectors extracted from the MPEG-1/2 bit-stream are integrated into the validation process thus providing a key source of information used in matching corresponding objects between frames. Intra-macroblock data are also extracted from the MPEG-1/2 bit-stream and is utilized in identifying new objects upon their entry into the scene. The algorithm is tested on a wide range of video sequences with results provided.

The objective of our work is to enable the optimum design of lightweight automotive structural components using injection-molded long fiber thermoplastics (LFTs). To this end, an integrated approach that links process modeling to structural analysis with experimental microstructural characterization and validation is developed. First, process models for LFTs are developed and implemented into processing codes (e.g. ORIENT, Moldflow) to predict the microstructure of the as-formed composite (i.e. fiber length and orientation distributions). In parallel, characterization and testing methods are developed to obtain necessary microstructural data to validate process modeling predictions. Second, the predicted LFT composite microstructure is imported into a structural finite element analysis by ABAQUS to determine the response of the as-formed composite to given boundary conditions. At this stage, constitutive models accounting for the composite microstructure are developed to predict various types of behaviors (i.e. thermoelastic, viscoelastic, elastic-plastic, damage, fatigue, and impact) of LFTs. Experimental methods are also developed to determine material parameters and to validate constitutive models. Such a process-linked-structural modeling approach allows an LFT composite structure to be designed with confidence through numerical simulations. Some recent results of our collaborative research will be illustrated to show the usefulness and applications of this integrated approach.

In subband/wavelet image coding, size-limited subband decompositions are ordinarily used to avoid increasing the number of samples that need to be coded. To reduce coding distortions that can occur at the borders, the symmetric extension filter bank is typically employed. This paper introduces some new perspectives and improvements to that decomposition. The symmetric extension filter bank is couched in the cyclic frequency domain, providing a framework that accommodates FIR and IIR filters in a natural way, all with perfect reconstruction. IIR filters with both rational and irrational transfer functions can be implemented and, in the context of symmetric extension, can accommodate IIRs that effectively have perfect stopband suppression. Enhancements to the filter bank at a tree-structured system level are also presented and include the application of spectral reversal correction and a transition band normalization approach to designing the constituent filters of the symmetric extension wavelet packet transform.

A new quality measurement for video sequences utilized in video retrieval systems and visual data mining applications is proposed. First, each frame of the sequence undergoes a segmentation step using extracted texture features from the gray-level cooccurrence matrix (GLCM) (Davis and Johns, 1979). Next, corresponding objects between adjacent frames are matched thus resulting in a 3-dimensional segmentation of the video into objects. Finally, color and texture features are extracted for each object in the sequence and provide the primary input in computing the quality measurement pertaining to the video. A low quality measurement may thus eliminate the possibility of the sequence being stored in a database retrieval system. The algorithm is tested on various types of video segments - pans, zooms, close-ups, and multiple objects' motion - with results included