Search

In recent years, electronic word of mouth (e-WOM) has been widely used by people around the world. Tripadvisor is an e-WOM travel website that provides information about reviews and opinions on travel-related content. To help users gather information faster, aspect-based sentiment analysis is necessary. Aspect-based sentiment analysis helps users to capture and extract important features from the reviews. Therefore, this study aims to build an aspect-based sentiment analysis model of Indonesian tourism review by extracting aspect-category and their corresponding polarities from user reviews. To gain the best model, we performed several experiments by using Convolutional Neural Networks (CNN). Moreover, we compared our CNN model with CNN-LSTM and CNN-GRU to identify the sentiment and aspects from the reviews. We also performed negation handling in our feature extraction process to improve our CNN models. Based on our experiments, CNN combined with both POS tag and negation handling outperformed the other models with the accuracy of sentiment analysis of 0.9522 and aspect category of 0.9551.

International audience; Camera relocalization is needed in several applications such as augmented reality or robot navigation. However, it is still challenging to have a both real-time and accurate method. In this paper, we present our hybrid method combing machine learning approach and geometric approach for real-time camera relocalization from a single RGB image. We introduce our sparse feature regression forest to improve the machine learning part. In our regression forest, we propose a novel split function, that uses a whole feature vector instead of classical binary test function to improve the accuracy of 2D-3D point correspondences. Moreover, we use sparse feature extraction (SURF features) to reduce time processing. The results indicate that our method is the only real-time hybrid method (50ms per frame). We also achieve results as accurate as the best state-of-the-art methods (hybrid methods) and outperform machine learning based and sparse feature based methods.; La relocalisation de caméra est nécessaire dans plusieurs applications telles que la réalité augmentée ou la navigation de robots. Cependant, il est encore difficile d'avoir une méthode à la fois précise et temps réel. Dans cet article, nous présentons notre méthode hybride combinant une approche d'apprentissage automatique et une approche géométrique pour la relocalisation de caméra en temps réel à partir d'une seule image RGB. Nous introduisons notre forêt de régression s'appuyant sur ces caractéristiques éparses pour améliorer la partie apprentissage automatique. Dans notre forêt de régression, nous proposons une nouvelle fonction de séparation qui utilise un vecteur caractéristique entier au lieu d'une fonction de test binaire classique pour améliorer la précision des correspondances 2D-3D des points. De plus, nous utilisons l'extraction de caractéristiques éparses (caractéristiques SURF) pour réduire le temps de traitement. Les résultats indiquent que notre méthode est la seule méthode hybride temps réel (50ms par image). Nous obtenons également des résultats aussi précis que les meilleures méthodes de l'état de l'art(méthodes hybrides) et surpassons les méthodes basées sur l'apprentissage automatique et les méthodes basées sur des caractéristiques éparses.

This paper deals with the design and simulation of a Split Ring Resonator (SRR) for the detection of leaks, corrosion and cracks in pipeline networks intended for transporting crude oil, natural gas, petroleum products and other fluid products. The designed SRR is tuned to a resonant frequency of 6.1 GHz and is simulated using High Frequency Structure Simulator (HFSS) tool to detect the ruptures in pipeline coatings made of FR4 epoxy. The designed SRR is mounted on a Rogers substrate above the FR4 epoxy coating and shows a notable variation in the resonant frequency and quality factor whenever gaps are detected between the coating and the pipeline, which can be used for checking the pipeline integrity.

Improvement in Science and Technology aims to better the world. The need for a comfortable and simplified life paves way to opportunities that can enhance the existing technology. The empowerment of differently abled will always stand as a priority. In this paper we propose the idea of a wristband that could convert sign (Hand gestures) to speech signals and enable the mute to enjoy the world in a better way. The wearable device uses the electrical impulses generated by movement of every finger to produce signs, each sign depicts a particular word. The band is a combination of wearable computer aided with sign to speech conversion techniques and a sensor suite. Compared to the existing methods, the proposed design provides a user-friendly experience and does not consider any specific parameter from the end user, this makes it a generalized band which can be worn by anybody. This would result in cost minimization when produced in large scale. The compactness of the design aims to provide more ease of usage.

To robustly estimate the pose, classical methods assume some geometrical and temporal assumptions (SfM: Structure from Motion, SLAM: Simultaneous Localization and mapping). These approaches take a pair of images as input and establish correspondences based on global strategy (using the whole image information) or sparse strategy (using key-points features). These correspondences allow solving a set of linear equations related to the 3D information and camera pose in that environment. These past years, machine learning has been considered as an efficient way to tackle different problems in image processing and computer vision fields. To handle the task in hand, we propose to learn directly the mapping function between the acquired information from the camera and its pose using sparse decision forest. We achieved state-of the-art results on public and on our databases.

When dealing with real estate business, it is quite difficult for estate agents to make customers understand the potential and the volumes of free spaces. Thus, we propose an application that aims to solve these issues based on a laying out scenario in which a seller and a customer collaborate. As the roles of both users are different, we propose an asymmetric collaboration where the two users do not use the same interaction setup and do not benefit from the same interaction capabilities.

International audience; Although compression of 3D content has been a widely studied topic, adaptive streaming of large virtual scenes is progressively becoming the standard way to share collaborative environments, due to the growing complexity of the shared data. The classical 'download and play' paradigm associated to standalone compression is leaving place to interactive refinements of regions-of-interest, provided given polygon budget and network capability. With the increasing number of applications and potential clients sharing reusable data, the traditional client / server architecture is likely to be over passed by less server-centered protocols, in order to distribute the complex tasks of identifying the selected data to be streamed. In this paper, we present the design and performances of a platform exploiting peer-to-peer transmission and show how the whole setting can be implemented using the MPEG-4 AFX standards.

Interactive network-based navigation over large urban environments raises difficult problems due to the size and complexity of these scenes. In this paper, we present a clientserver system allowing navigation over 3D cities in real time. Due to a novel progressive and hierarchical representation of 3D models of densely built urban areas, only perceptible details for all the regions visible from a given viewpoint are progressively streamed to visualisation clients. Furthermore, efficient coding methods are used to compress the representation data allowing quick start-up of the interactive visualisation with a highly-detailed model. This is achieved through a set of dedicated algorithms allowing a very large city model to be structured into a multi-resolution representation. The method efficiently exploits the fact that most automated modelling techniques of urban scenes provides 2D\frac{1} {2} models (building footprint, height, altitude, ...). So as to efficiently and faithfully model complex buildings, a procedural representation for roofs and facades is proposed. Finally, we present an MPEG4 compatible implementation based on the introduction of new node types with the associated bitstream.