Structure Preserving Non-Photorealistic Rendering Framework for Image Abstraction and Stylization of Low-Illuminated and Underexposed Images

Author(s):  
Pavan Kumar ◽  
Poornima B. ◽  
Nagendraswamy H. S. ◽  
Manjunath C.

The proposed abstraction framework manipulates the visual-features from low-illuminated and underexposed images while retaining the prominent structural, medium scale details, tonal information, and suppresses the superfluous details like noise, complexity, and irregular gradient. The significant image features are refined at every stage of the work by comprehensively integrating a series of AnshuTMO and NPR filters through rigorous experiments. The work effectively preserves the structural features in the foreground of an image and diminishes the background content of an image. Effectiveness of the work has been validated by conducting experiments on the standard datasets such as Mould, Wang, and many other interesting datasets and the obtained results are compared with similar contemporary work cited in the literature. In addition, user visual feedback and the quality assessment techniques were used to evaluate the work. Image abstraction and stylization applications, constraints, challenges, and future work in the fields of NPR domain are also envisaged in this paper.

2021 ◽  
Vol 13 (1) ◽  
pp. 1-35
Author(s):  
Pavan Kumar ◽  
Poornima B. ◽  
H. S. Nagendraswamy ◽  
C. Manjunath ◽  
B. E. Rangaswamy

Underexposed heterogeneous complex-background and graphical embossing text documents are treated using proposed preprocessing image-abstraction framework that can deliver the effective structure preserved abstracted output by manipulating visual-features from input images. Reading of the text character in such images is extremely poor; hence, the framework effectively boosted the significant image properties and quality features at every stage. Work effectively preserves the foreground structure of an image by comprehensively integrating the sequence of NPR filters and diminishes the background content of an image, and in this way, the framework contributes to separation of foreground text from image background. Effectiveness of the proposed work has been validated by conducting the trials on the selected dataset. In addition, user's visual-feedback and image quality assessment techniques were also used to evaluate the framework. Based on the obtained abstraction output, this work extracts text-character by wisely utilizing traditional image processing techniques with an average accuracy of 98.91%.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Abdulkadir Tasdelen ◽  
Baha Sen

AbstractmiRNAs (or microRNAs) are small, endogenous, and noncoding RNAs construct of about 22 nucleotides. Cumulative evidence from biological experiments shows that miRNAs play a fundamental and important role in various biological processes. Therefore, the classification of miRNA is a critical problem in computational biology. Due to the short length of mature miRNAs, many researchers are working on precursor miRNAs (pre-miRNAs) with longer sequences and more structural features. Pre-miRNAs can be divided into two groups as mirtrons and canonical miRNAs in terms of biogenesis differences. Compared to mirtrons, canonical miRNAs are more conserved and easier to be identified. Many existing pre-miRNA classification methods rely on manual feature extraction. Moreover, these methods focus on either sequential structure or spatial structure of pre-miRNAs. To overcome the limitations of previous models, we propose a nucleotide-level hybrid deep learning method based on a CNN and LSTM network together. The prediction resulted in 0.943 (%95 CI ± 0.014) accuracy, 0.935 (%95 CI ± 0.016) sensitivity, 0.948 (%95 CI ± 0.029) specificity, 0.925 (%95 CI ± 0.016) F1 Score and 0.880 (%95 CI ± 0.028) Matthews Correlation Coefficient. When compared to the closest results, our proposed method revealed the best results for Acc., F1 Score, MCC. These were 2.51%, 1.00%, and 2.43% higher than the closest ones, respectively. The mean of sensitivity ranked first like Linear Discriminant Analysis. The results indicate that the hybrid CNN and LSTM networks can be employed to achieve better performance for pre-miRNA classification. In future work, we study on investigation of new classification models that deliver better performance in terms of all the evaluation criteria.


Author(s):  
M. P. Pavan Kumar ◽  
B. Poornima ◽  
H. S. Nagendraswamy ◽  
C. Manjunath

Author(s):  
Anne H.H. Ngu ◽  
Jialie Shen ◽  
John Shepherd

The optimized distance-based access methods currently available for multimedia databases are based on two major assumptions: a suitable distance function is known a priori, and the dimensionality of image features is low. The standard approach to building image databases is to represent images via vectors based on low-level visual features and make retrieval based on these vectors. However, due to the large gap between the semantic notions and low-level visual content, it is extremely difficult to define a distance function that accurately captures the similarity of images as perceived by humans. Furthermore, popular dimension reduction methods suffer from either the inability to capture the nonlinear correlations among raw data or very expensive training cost. To address the problems, in this chapter we introduce a new indexing technique called Combining Multiple Visual Features (CMVF) that integrates multiple visual features to get better query effectiveness. Our approach is able to produce low-dimensional image feature vectors that include not only low-level visual properties but also high-level semantic properties. The hybrid architecture can produce feature vectors that capture the salient properties of images yet are small enough to allow the use of existing high-dimensional indexing methods to provide efficient and effective retrieval.


Electronics ◽  
2020 ◽  
Vol 9 (12) ◽  
pp. 2162
Author(s):  
Changqi Sun ◽  
Cong Zhang ◽  
Naixue Xiong

Infrared and visible image fusion technologies make full use of different image features obtained by different sensors, retain complementary information of the source images during the fusion process, and use redundant information to improve the credibility of the fusion image. In recent years, many researchers have used deep learning methods (DL) to explore the field of image fusion and found that applying DL has improved the time-consuming efficiency of the model and the fusion effect. However, DL includes many branches, and there is currently no detailed investigation of deep learning methods in image fusion. In this work, this survey reports on the development of image fusion algorithms based on deep learning in recent years. Specifically, this paper first conducts a detailed investigation on the fusion method of infrared and visible images based on deep learning, compares the existing fusion algorithms qualitatively and quantitatively with the existing fusion quality indicators, and discusses various fusions. The main contribution, advantages, and disadvantages of the algorithm. Finally, the research status of infrared and visible image fusion is summarized, and future work has prospected. This research can help us realize many image fusion methods in recent years and lay the foundation for future research work.


2021 ◽  
Vol 18 (180) ◽  
pp. 20210324
Author(s):  
Karl T. Bates ◽  
Linjie Wang ◽  
Matthew Dempsey ◽  
Sarah Broyde ◽  
Michael J. Fagan ◽  
...  

Measures of attachment or accommodation area on the skeleton are a popular means of rapidly generating estimates of muscle proportions and functional performance for use in large-scale macroevolutionary studies. Herein, we provide the first evaluation of the accuracy of these muscle area assessment (MAA) techniques for estimating muscle proportions, force outputs and bone loading in a comparative macroevolutionary context using the rodent masticatory system as a case study. We find that MAA approaches perform poorly, yielding large absolute errors in muscle properties, bite force and particularly bone stress. Perhaps more fundamentally, these methods regularly fail to correctly capture many qualitative differences between rodent morphotypes, particularly in stress patterns in finite-element models. Our findings cast doubts on the validity of these approaches as means to provide input data for biomechanical models applied to understand functional transitions in the fossil record, and perhaps even in taxon-rich statistical models that examine broad-scale macroevolutionary patterns. We suggest that future work should go back to the bones to test if correlations between attachment area and muscle size within homologous muscles across a large number of species yield strong predictive relationships that could be used to deliver more accurate predictions for macroevolutionary and functional studies.


Author(s):  
Lynn Randall ◽  
Pierre Zundel

There have been calls in the literature for reforms to assessment to enhance student learning (Shepard, 2000). In many instances, this refers to the need to move from traditional assessment procedures that are characterized as content-heavy, summative, and norm-referenced approaches to more constructivist and student-centred approaches, often characterized as more “…flexible, integrative, contextualized, process oriented, criteria referenced and formative” (Ellery, 2008, p. 421). Whereas summative assessment techniques rarely allow students to act on the feedback provided, formative feedback provided throughout the learning process can be used to improve future work and promote learning (Ellery, 2008; Higgins, Hartley & Skelton, 2002) by providing students an opportunity to learn from mistakes. Allowing students to learn from their mistakes makes good pedagogical sense. To date there has been little research examining students’ use of feedback (Higgins, Hartley, and Skelton, 2002). In an effort to begin to add to the literature in this area, this paper describes a study that explored the effectiveness of oral and written formative feedback when students were provided the opportunity to use it. The paper begins by reviewing literature related to assessment and how assessment relates to feedback in general. It then presents what the research has found in relation to students’ perspectives of effective feedback and how they use it. The paper ends by presenting the results and discussion. La documentation fait état de demandes de réforme de l’évaluation pour améliorer l’apprentissage des étudiants (Shepard, 2000). Dans plusieurs cas, cela traduit le besoin de passer des procédures d’évaluation traditionnelles caractérisées par la lourdeur de leur contenu, par leur aspect sommatif et par leurs approches normatives à des approches plus constructivistes et centrées sur les étudiants, souvent qualifiées de plus « ... souples, intégratives, contextualisées, axées sur les processus, balisées par des critères et formatives » (Ellery, 2008, p. 421). Alors que les techniques d’évaluation sommative permettent rarement aux étudiants de se conformer à la rétroaction fournie, la rétroaction formative tout au long du processus d’apprentissage peut être utilisée pour améliorer les travaux futurs et favoriser l’apprentissage (Ellery, 2008; Higgins, Hartley et Skelton, 2002) en donnant l’occasion aux étudiants d’apprendre de leurs erreurs. Sur le plan pédagogique, permettre aux étudiants d’apprendre de leurs erreurs a du sens. À ce jour, il y a eu peu de recherche sur l’utilisation que font les étudiants de la rétroaction (Higgins, Hartley et Skelton, 2002). Le présent article se veut un ajout à la documentation dans ce domaine. Ses auteurs décrivent une étude qui porte sur l’efficacité de la rétroaction formative orale et écrite lorsque les étudiants ont eu l’occasion de l’utiliser subséquemment. Les auteurs commencent par une analyse de la documentation sur l’évaluation et sur les liens généraux entre cette dernière et la rétroaction. Ils présentent ensuite les résultats de recherche liée aux perspectives des étudiants sur la rétroaction efficace et sur l’utilisation qu’ils en font. Enfin, ils terminent par une présentation des résultats et par une discussion.


2018 ◽  
Vol 1 (1) ◽  
Author(s):  
Jodee Jacobs ◽  
Cristie McClendon

The purpose of this article was to explore the use of reflective practice in teaching in two online doctoral research courses. Gibbs’ (1988) model was used as a framework for reflection. The steps from this model are centered on description, feelings, evaluation, analysis, conclusions, and action plans to enrich one’s practice. Two faculty members collaborated to discuss concerns related to doctoral learners’ struggles with being able to identify and evaluate key components of research studies. Results from reflections indicated the faculty identified that students held some responsibility for self-direction and completing their assignments. However, faculty identified that three areas could be improved upon in future courses. Assignment instructions need to be further clarified. Scaffolding strategies such as classroom assessment techniques, exemplar papers, Zoom conferences, and clarifying questions are necessary to facilitate student success. Finally, more focused feedback specific to the assignments for students to use on future work is necessary. Recommendations for future research include a study of learners’ perspectives of how they use feedback to learn how to identify research study components, chow use of classroom assessment techniques can be used to scaffold instruction in online courses where the materials are preloaded, and differences between individual and team reflective practices. Keywords: Reflective practice, Gibbs Model, Feedback, Scaffolding, Assignments


Geophysics ◽  
2020 ◽  
Vol 85 (4) ◽  
pp. WA27-WA39 ◽  
Author(s):  
Xinming Wu ◽  
Zhicheng Geng ◽  
Yunzhi Shi ◽  
Nam Pham ◽  
Sergey Fomel ◽  
...  

Seismic structural interpretation involves highlighting and extracting faults and horizons that are apparent as geometric features in a seismic image. Although seismic image processing methods have been proposed to automate fault and horizon interpretation, each of which today still requires significant human effort. We improve automatic structural interpretation in seismic images by using convolutional neural networks (CNNs) that recently have shown excellent performances in detecting and extracting useful image features and objects. The main limitation of applying CNNs in seismic interpretation is the preparation of many training data sets and especially the corresponding geologic labels. Manually labeling geologic features in a seismic image is highly time-consuming and subjective, which often results in incompletely or inaccurately labeled training images. To solve this problem, we have developed a workflow to automatically build diverse structure models with realistic folding and faulting features. In this workflow, with some assumptions about typical folding and faulting patterns, we simulate structural features in a 3D model by using a set of parameters. By randomly choosing the parameters from some predefined ranges, we are able to automatically generate numerous structure models with realistic and diverse structural features. Based on these structure models with known structural information, we further automatically create numerous synthetic seismic images and the corresponding ground truth of structural labels to train CNNs for structural interpretation in field seismic images. Accurate results of structural interpretation in multiple field seismic images indicate that our workflow simulates realistic and generalized structure models from which the CNNs effectively learn to recognize real structures in field images.


2014 ◽  
Vol 4 (1) ◽  
pp. 37-59 ◽  
Author(s):  
Shawn Conrad ◽  
Jody Clarke-Midura ◽  
Eric Klopfer

Educational games offer an opportunity to engage and inspire students to take interest in science, technology, engineering, and mathematical (STEM) subjects. Unobtrusive learning assessment techniques coupled with machine learning algorithms can be utilized to record students' in-game actions and formulate a model of the students' knowledge without interrupting the students' play. This paper introduces “Experiment Centered Assessment Design” (XCD), a framework for structuring a learning assessment feedback loop. XCD builds on the “Evidence Centered Assessment Design” (ECD) approach, which uses tasks to elicit evidence about students and their learning. XCD defines every task as an experiment in the scientific method, where an experiment maps a test of factors to observable outcomes. This XCD framework was applied to prototype quests in a massively multiplayer online (MMO) educational game. Future work would build upon the XCD framework and use machine learning techniques to provide feedback to students, teachers, and researchers.


Sign in / Sign up

Export Citation Format

Share Document