Our work focuses on detecting sarcasm in tweets using deep learning extracted features combined with contextual handcrafted features. A feature set is from a Convolutional Neural Network (CNN) architecture before it carefully sets. These sets are created based their respective explanations. Each specifically designed for the sole task of detection. The objective to find most optimal Some good go even when used independence. Other not really significant without any combination. results...

Urban screens are becoming a common element of our city landscape. As such they offer new ways connecting people that occupy public space, e.g., by taking situated snapshots through display-attached camera. In this paper we present first longitudinal case study 12 weeks an application -- Moment Machine deployed "in the wild" on urban screen facing street. We report findings from 1189 photos taken, 13 interviews, and 3 observations show engagement stimulated within place-based community where...

High-quality development is the theme of China’s economic and social in new era, only way to build a powerful tourism economy, also objective demand solve practical problems promote transformation upgrading tourism. However, such as inefficient operation, environmental pollution, slow industrial have become prominent process industry recent years, which gap exists with realization high-quality goals. Therefore, scientific control deviation between actual level goals helpful for promoting...

Efficient obstacle-avoidance path planning is critical for orchards with numerous irregular obstacles. This paper presents a continuous bidirectional Quick-RRT* (CBQ-RRT*) algorithm based on the RRT (Bi-RRT) and algorithms proposes an expansion cost function that evaluates smoothness length to overcome limitations of non-holonomic mobile robot applications. To improve zigzag between dual trees caused by dual-tree Bi-RRT algorithm, CBQ-RRT* CreateConnectNode optimization method, which...

Generating legal and diverse layout patterns to establish large pattern libraries is fundamental for many lithography design applications. Existing generation models typically regard the problem as image of maps learn model via capturing pixel-level coherence, which insufficient achieve polygon-level modeling, e.g., shape patterns, thus leading poor quality. In this paper, we an unsupervised sequence problem, in order rules by explicitly modeling shapes polygons layouts among polygons....

With the rapid development of Web services technologies, discovering is becoming most urgent problem to be resolved. The existing service discovery methods are not mature, and cannot cope with requests from different fields very well. In this paper we suggest an effective feasible approach based on functional semantics. We define semantic description that provides a unified manner semantically describe function for providers customers. order avoid heterogeneity, build domain-oriented...

In this paper, we describe our Research-in-the-Wild driven methodology to tackle a complex range of social, technical and interactional issues when implementing networked Urban Screens in London Nottingham, UK [1]. The diversity the local communities provides unique opportunity for research examine interactions within town centres, as well UK-wide.

Catheters and wires are used extensively in cardiac catheterization procedures. Detecting their positions fluoroscopic X-ray images is important for several clinical applications such as motion compensation co-registration between 2D 3D imaging modalities. the complete length of a catheter or wire object well electrode on challenging task.In this paper, an automatic detection framework catheters developed. It based path reconstruction from image tensors, which eigen direction vectors...

Automatic image content annotation techniques attempt to explore structural visual features of images that describe and associate them with semantics. In this paper, two types concept spaces, atomic collective are defined the problems in those spaces formulated as feature classification Bayesian inference, respectively. A scheme framework is presented evaluated an application photo categorization using MPEG-7 VCE2 dataset its ground truth. The experimental results show a promising performance

Networked public displays are a new medium, and as such they can improve the quality experience of urban spaces by providing means to engage members place-based communities. As located in different understanding commonalities differences between their users helps building solid foundation for knowledge on how networked operate. In this paper, we describe findings from 12 weeks deployment Moment Machine application that allows posting viewing situated snapshots across 4 placed settings...

Although they have achieved great success in optical images, deep convolutional neural networks underperform for ship detection SAR images because of the lack color and textual features. In this paper, we propose our framework which integrates prior knowledge into by means attention mechanism. Because background ships is mostly water surface or coast, use clustering algorithms to generate map from brightness density The later resized fused with feature maps Our experiments demonstrate that...

The secure verification is important for watermarking protocols. A malicious arbitrator able to remove an original watermark from unauthorized copy of the digital content as a result security breach in phase arbitration and resell multiple copies it with impunity. We propose novel buyer-seller protocol verification. In this scheme, seller permutes provided by trusted Watermarking Certification Authority (WCA) embeds into encrypted domain. case found, can recover extracted sends arbitrator....

Conventional relevance feedback schemes may not be suitable to all practical applications of content-based image retrieval (CBIR), since most ordinary users would like complete their search in a single interaction, especially on the web search. In this paper, we explore new approach improve performance based concept, bag images, rather than feedback. We consider that collection comprises bags instead independent individual images. Each includes some relevant images with same perceptual...

Content based image retrieval (CBIR) is a technique to search for images relevant the user's query from an collection. In last decade, most attention has been paid improve performance. However, there no significant effort investigate security concerning in CBIR. Under by example (QBE) paradigm, user supplies as and system returns set of retrieved results. If includes private information, untrusted server provider CBIR may distribute it illegally, which leads right problem. this paper, we...