13,209 research outputs found
Author Correction: Evaluation of skin cancer resection guide using hyper‑realistic in‑vitro phantom fabricated by 3D printing
No full textThe original version of this Article contained an error in the spelling of the author Taehun Kim which was incorrectly given as Teahun Kim. The original Article has been corrected
Methodology for extracting the delighter in Kano model using big data analysis
No full textWith shortening product lifecycles, product design is more strongly influencing successful business’ competitive advantage. The Kano model has been proposed to define and extract customers’ needs for attractive quality creation in product development. Because product or service differentiation is crucial to business successes, the extraction of the delighter in the Kano model is an important issue. This study proposes a methodology for extracting important factors by using big data outside the organisation; such a factor may not be a megatrend but may be a delighter in the Kano model. Because the delighter in the Kano model is defined as a constant necessity arising from a small number of propounders, the volume concept is used by adding the time axis to the flatland of the data distribution to extract the delighter. To demonstrate the feasibility of the proposed methodology, accumulated data from the iPhoneForum from 2010 are gathered and analysed to extract the delighter for a smartphone’s input device. The result shows that the ‘pen’ and ‘write’ have the highest index of potential delighter during the study period; in particular, before December 2014, ‘pen’ had the highest index of potential delighter.11Nssciscopu
SpaceMeshLab: Spatial Context Memoization And Meshgrid Atrous Convolution Consensus For Semantic Segmentation
No full text1
Spatio-Temporal Slowfast Self-Attention Network For Action Recognition
No full textWe propose Spatio-Temporal SlowFast Self-Attention network for action recognition. Conventional Convolutional Neural Networks have the advantage of capturing the local area of the data. However, to understand a human action, it is appropriate to consider both human and the overall context of given scene. Therefore, we repurpose a self-attention mechanism from Self-Attention GAN (SAGAN) to our model for retrieving global semantic context when making action recognition. Using the self-attention mechanism, we propose a module that can extract four features in video information: spatial information, temporal information, slow action information, and fast action information. We train and test our network on the Atomic Visual Actions (AVA) dataset and show significant frame-AP improvements on 28 categories.1
Metadata-oriented Methodology for Building a Data Warehouse: a Medical Center Case
No full textData warehouse is an intelligent store of data that can manage and aggregate vast amounts of information. A metadata is critical for implementing data warehouse. Therefore, integrating data warehouse with its metadata offers a new opportunity to create a more adaptive and flexible information system. This paper proposes a metadata-oriented methodology for building data warehouse that consists of seven components: legacy, extracting operational data store, data warehouse, data mart, application, and metadata. A taxonomy for dataflow and metaflow is proposed for better understanding of the methodology. In addition, a metadata schema is built within the framework of the seven components. The methodology by using its metadata component is applied to real-life data warehouse for a large medical center in order to illustrate its practical usefulness
Semi-Supervised 3D Object Detection With Channel Augmentation Using Transformation Equivariance
No full textUACANet: Uncertainty Augmented Context Attention for Polyp Segmentation
No full textWe propose Uncertainty Augmented Context Attention network (UACANet) for polyp segmentation which considers an uncertain area of the saliency map. We construct a modified version of U-Net shape network with additional encoder and decoder and compute a saliency map in each bottom-up stream prediction module and propagate to the next prediction module. In each prediction module, previously predicted saliency map is utilized to compute foreground, background and uncertain area map and we aggregate the feature map with three area maps for each representation. Then we compute the relation between each representation and each pixel in the feature map. We conduct experiments on five popular polyp segmentation benchmarks, Kvasir, CVC-ClinicDB, ETIS, CVC-ColonDB and CVC-300, and our method achieves state-of-the-art performance. Especially, we achieve 76.6% mean Dice on ETIS dataset which is 13.8% improvement compared to the previous state-of-the-art method. Source code is publicly available at https://github.com/plemeri/UACANet1
- …
