FineLocator: A novel approach to method-level fine-grained bug localization by query expansion

2019 ◽  
Vol 110 ◽  
pp. 121-135 ◽  
Author(s):  
Wen Zhang ◽  
Ziqiang Li ◽  
Qing Wang ◽  
Juan Li
Keyword(s):  
Query Expansion ◽  
Bug Localization ◽  
Fine Grained ◽  
Novel Approach

scholarly journals Explaining Neural Networks Using Attentive Knowledge Distillation

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1280
Author(s):  
Hyeonseok Lee ◽  
Sungchan Kim
Keyword(s):  
Neural Networks ◽  
Model Prediction ◽  
Saliency Map ◽  
Model Parameters ◽  
Learning Capability ◽  
Fine Grained ◽  
Network Layers ◽  
Saliency Maps ◽  
Novel Approach ◽  
Knowledge Distillation

Explaining the prediction of deep neural networks makes the networks more understandable and trusted, leading to their use in various mission critical tasks. Recent progress in the learning capability of networks has primarily been due to the enormous number of model parameters, so that it is usually hard to interpret their operations, as opposed to classical white-box models. For this purpose, generating saliency maps is a popular approach to identify the important input features used for the model prediction. Existing explanation methods typically only use the output of the last convolution layer of the model to generate a saliency map, lacking the information included in intermediate layers. Thus, the corresponding explanations are coarse and result in limited accuracy. Although the accuracy can be improved by iteratively developing a saliency map, this is too time-consuming and is thus impractical. To address these problems, we proposed a novel approach to explain the model prediction by developing an attentive surrogate network using the knowledge distillation. The surrogate network aims to generate a fine-grained saliency map corresponding to the model prediction using meaningful regional information presented over all network layers. Experiments demonstrated that the saliency maps are the result of spatially attentive features learned from the distillation. Thus, they are useful for fine-grained classification tasks. Moreover, the proposed method runs at the rate of 24.3 frames per second, which is much faster than the existing methods by orders of magnitude.

Modeling Secure 3D Web Applications

2011 ◽  
pp. 263-283 ◽  
Author(s):  
Krzysztof Walczak
Keyword(s):  
Web Applications ◽  
Fine Grained ◽  
Dynamic Composition ◽  
Novel Approach ◽  
Access Privileges ◽  
3D Scene ◽  
High Level ◽  
And Behavior ◽  
Content Modeling ◽  
3D Web

This chapter describes a novel approach to building 3D web applications, called Flex-VR, which can be used a basis for implementing security solutions. Two key elements of the approach are described: scene structuralization and content modeling. The scene structuralization enables decomposition of a 3D scene into independent geometrical and behavioral objects, called VR-Beans. Virtual scenes with rich interactivity and behavior can be dynamically created by combining sets of independent VR-Beans. The second element – the content model – is a generalized high-level description of the application content. The model enables efficient manipulation of content elements and dynamic composition of virtual scenes. Flex-VR provides a fine-grained semantically-rich content structure, which can be used as a basis for defining access privileges for users and groups. Five levels of user privileges definition in the Flex-VR approach are described. An application of Flex-VR in the cultural heritage domain is presented. Examples demonstrate how user privileges can be defined at all levels.

Inducing and Refining Topics for Web Query Classification Using a Semantic Network

2018 ◽  
Vol 15 (2) ◽  
pp. 595-600
Author(s):  
R. Sathish Kumar ◽  
M. Chandrasekaran
Keyword(s):  
Query Expansion ◽  
Web Search ◽  
Semantic Network ◽  
Machine Learning Techniques ◽  
Search Query ◽  
Fine Grained ◽  
Computational Overhead ◽  
Learning Techniques ◽  
Query Classification ◽  
The Web

Web query classification, the task of inferring topical categories from a web search query is a non-trivial problem in Information Retrieval domain. The topic categories inferred by a Web query classification system may provide a rich set of features for improving query expansion and web advertising. Conventional methods for Web query classification derive corpus statistics from the web and employ machine-learning techniques to infer Open Directory Project categories. But they suffer from two major drawbacks, the computational overhead to derive corpus statistics and inferring topic categories that are too abstract for semantic discrimination due to polysemy. Concepts too shallow or too deep in the semantic gradient are produced due to the wrong senses of the query terms coalescing with the correct senses. This paper proposes and demonstrates a succinct solution to these problems through a method based on the Tree cut model and Wordnet Thesarus to infer fine-grained topic categories for Web query classification, and also suggests an enhancement to the Tree Cut Model to resolve sense ambiguities.

scholarly journals Microstructure refinement-homogenization and flexural strength improvement of Al2O3 ceramics fabricated by DLP-stereolithography integrated with chemical precipitation coating process

2021 ◽  
Author(s):  
Guanglin Nie ◽  
Yehua Li ◽  
Pengfei Sheng ◽  
Fei Zuo ◽  
Haolin Wu ◽  
...  
Keyword(s):  
Flexural Strength ◽  
Chemical Precipitation ◽  
Heat Sinks ◽  
Coarse Grained ◽  
Homogeneous Layer ◽  
Mechanical Mixing ◽  
Fine Grained ◽  
Novel Approach ◽  
Ceramic Components ◽  
3D Printed

AbstractIn this study, the chemical precipitation coating (CP) process was creatively integrated with DLP-stereolithography based 3D printing for refining and homogenizing the microstructure of 3D printed Al2O3 ceramic. Based on this novel approach, Al2O3 powder was coated with a homogeneous layer of amorphous Y2O3, with the coated Al2O3 powder found to make the microstructure of 3D printed Al2O3 ceramic more uniform and refined, as compared with the conventional mechanical mixing (MM) of Al2O3 and Y2O3 powders. The grain size of Al2O3 in Sample CP is 64.44% and 51.43% lower than those in the monolithic Al2O3 ceramic and Sample MM, respectively. Sample CP has the highest flexural strength of 455.37±32.17 MPa, which is 14.85% and 25.45% higher than those of Samples MM and AL, respectively; also Sample CP has the highest Weibull modulus of 16.88 among the three kinds of samples. Moreover, the fine grained Sample CP has a close thermal conductivity to the coarse grained Sample MM because of the changes in morphology of Y3Al5O12 phase from semi-connected (Sample MM) to isolated (Sample CP). Finally, specially designed fin-type Al2O3 ceramic heat sinks were successfully fabricated via the novel integrated process, which has been proven to be an effective method for fabricating complex-shaped Al2O3 ceramic components with enhanced flexural strength and reliability.

scholarly journals Novel Approach for Fine Ilmenite Flotation Using Hydrophobized Glass Bubbles as the Buoyant Carrier

Minerals ◽  
2021 ◽  
Vol 11 (3) ◽  
pp. 231
Author(s):  
Pan Chen ◽  
Youchuan Chen ◽  
Hang Liu ◽  
Haoyu Li ◽  
Xujian Chai ◽  
...  
Keyword(s):  
Froth Flotation ◽  
Interaction Force ◽  
Optical Microscope ◽  
Dlvo Theory ◽  
Microscope Observation ◽  
Base Interaction ◽  
Simple Modification ◽  
Fine Grained ◽  
Novel Approach ◽  
Hydrophobic Particles

Ilmenite disseminated grain size is relatively fine, and it must be finely ground to fully separate ilmenite from gangue and then produce fine-grained minerals, which deteriorates flotation. A novel method using buoyant carriers to improve the recovery of fine ilmenite in froth flotation was introduced in this study. Hydrophobized glass bubbles (HGB) as carrier materials were obtained by an efficient, simple modification of ordinary glass bubbles. The carrier flotation of fine ilmenite in the presence of HGB was investigated by micro flotation tests, X-ray diffractometer analysis, Fourier transform infrared (FTIR), optical microscope observation, and the extended DLVO theory (XDLVO). Micro-flotation results showed that the recovery of fine ilmenite in presence of HGB was 37.7% higher than that when using NaOL alone at pH 6. FTIR analysis and optical microscope observation revealed that fine ilmenite particles can be closely attached on the HGB surface to increase apparent particle size considerably. The data calculated from the DLVO theory indicated that the acid–base interaction force determined the adsorption between two hydrophobic particles.

scholarly journals KL-MOB Automated Covid-19 Recognition Using a Novel Approach Based on Image Enhancement and a Modified MobileNet CNN

2021 ◽  
Author(s):  
Mundher Mohammed Taresh ◽  
Ning bo Zhu ◽  
Asaad Shakir Hameed ◽  
Modhi Lafta Mutar ◽  
Talal Ahmed Ali Ali Ahmed Ali Ali ◽  
...  
Keyword(s):  
Neural Network ◽  
Loss Function ◽  
X Ray ◽  
Kl Divergence ◽  
Fine Grained ◽  
Novel Approach ◽  
Starting Point ◽  
The World ◽  
Novel Coronavirus ◽  
Sensitivity Specificity

The emergence of the novel coronavirus pneumonia (Covid-19) pandemic at the end of 2019 led to chaos worldwide. The world breathed a sigh of relief when some countries announced that they had obtained the appropriate vaccine and gradually began to distribute it. Nevertheless, the emergence of another wave of this disease has returned us to the starting point. At present, early detection of infected cases has been the paramount concern of both specialists and health researchers. This paper aims to detect infected patients through chest X-ray images. The large dataset available online for Covid-19 (COVIDx) was used in this research. The dataset consists of 2,128 x-ray images of Covid-19 cases, 8,066 normal cases, and 5,575 cases of pneumonia. A hybrid algorithm was applied to improve image quality before conducting the neural network training process. This algorithm consisted of combining two different noise reduction filters in the images, followed by a contrast enhancement algorithm. In this paper, for Covid-19 detection, a novel convolution neural network (CNN) architecture, KL-MOB (Covid-19 detection network based on MobileNet structure), was proposed. KL-MOB performance was boosted by adding the Kullback Leibler (KL) divergence loss function at the end when trained from scratch. The Kullback-Leibler (KL) divergence loss function was adopted as content-based image retrieval and fine-grained classification to improve the quality of image representation. This paper yielded impressive results, overall benchmark accuracy, sensitivity, specificity, and precision of 98.7%, 98.32%, 98.82%, and 98.37%, respectively. The promising results in this research may enable other researchers to develop modern and innovative methods to aid specialists. The tremendous potential of the method proposed in this research can also be utilized to detect Covid-19 quickly and safely in patients throughout the world.

scholarly journals Affective Multimodal Control of Virtual Characters

2020 ◽  
Vol 6 (4) ◽  
pp. 43-54 ◽  
Author(s):  
Martin Klesen ◽  
Patrick Gebhard
Keyword(s):  
Real Time ◽  
The Body ◽  
Temporal Constraints ◽  
Virtual Characters ◽  
Fine Grained ◽  
Novel Approach ◽  
Control Body ◽  
High Level ◽  
The Mind ◽  
Selection Of

In this paper we report about the use of computer generated affect to control body and mind of cognitively modeled virtual characters. We use the computational model of affect ALMA that is able to simulate three different affect types in real-time. The computation of affect is based on a novel approach of an appraisal language. Both the use of elements of the appraisal language and the simulation of different affect types has been evaluated. Affect is used to control facial expressions, facial complexions, affective animations, posture, and idle behavior on the body layer and the selection of dialogue strategies on the mind layer. To enable a fine-grained control of these aspects a Player Markup Language (PML) has been developed. The PML is player-independent and allows a sophisticated control of character actions coordinated by high-level temporal constraints. An Action Encoder module maps the output of ALMA to PML actions using affect display rules. These actions drive the real-time rendering of affect, gesture and speech parameters of virtual characters, which we call Virtual Humans. 

Knowledge creation in patent ecosystems: insights from Singapore

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Carmela Elita Schillaci ◽  
Elona Marku ◽  
Manuel Castriotta ◽  
Maria Chiara Di Guardo
Keyword(s):  
High Tech ◽  
The Novel ◽  
Content Type ◽  
Government Interventions ◽  
Codified Knowledge ◽  
Knowledge Domains ◽  
Fine Grained ◽  
Novel Approach ◽  
Ecosystem Shift ◽  
Ecosystem Level

Purpose This paper aims to better understand how codified knowledge that originates in organizations contributes to the generation of idiosyncratic knowledge embedded at a more expansive level, such as that of an ecosystem. In doing so, the authors introduce the concept of patent ecosystems – conceived as configurations of codified knowledge advancements protected via patents. Design/methodology/approach Using a patent co-classification method and introducing a novel validated software, the authors map and visualize the patent ecosystem of Singapore and examine 173,597 patents published from 1995 to 2020. Findings Results reveal the prominent growth of Singapore’s patenting activities, capturing a patent ecosystem shift, from a more diverse knowledge configuration to a more specialized one. The codified knowledge mainly generated deals with pharmaceuticals and high-tech knowledge domains; further, newly emerging technologies such as blockchain are also noted. Research limitations/implications The research investigates Singapore’s context, a country in which research directions and focus areas are influenced by government interventions and leadership. Thus, future studies might examine other patent ecosystems to draw comparisons with more laissez-faire policies or ecosystems with more pronounced organic development. Originality/value The novelty of this research is the introduction of the concept of a patent ecosystem for advancing a more fine-grained understanding of the aggregated knowledge generated at the ecosystem level and its specific features, composition and development. The authors consider patents as “carriers” of different codified pieces of knowledge and patent ecosystems represent the configuration that emerges from connections of these elements. The novel approach can aid both researchers, practitioners and policymakers with future examinations in the field.

Sign in / Sign up

Export Citation Format

Share Document