Citation Context

...sors are incessantly active while most of others are activated occasionally, which causes the long-tail phenomenon. A representative example of social sensing is the construction of indoor floorplans =-=[1, 28]-=-. This research topic has recently drawn a growing interest since it potentially can support a wide range of location-based applications. The goal is to develop an automatic floorplan construction sys...

Citation Context

...nments. For example, the pathway areas usually have much denser training data than the inner parts of the shops in a mall. There are many reasons for the surveyor cannot perform long time data collection inside shops without getting consent from the shop owner; it is more challenging to figure out precise positions within shops than on the pathways as the maps contain little information about inner shops; pathways can be surveyed by regular walking while inside shops have to be collected point by point. The situation will not change substantially for crowdsourcing-based data collection (e.g., [22, 31, 33, 38]) as popular places and pathways will always have more chances to be visited and hence enjoy more frequent data collection. In addition, the WiFi infrastructure may be updated with APs removed, relocated or added. Thus, it is unlikely to obtain a location database that has all indoor areas sampled dense enough and up to date. Environmental Impact: Indoor environments are typically divided into many smaller functional areas such as corridors, rooms and shops, among others. They are usually separated using cement, wood or glass walls that cause different penetrations and reflections of radio sig...

Citation Context

...ited the gyroscope for heading direction estimation. Their reports also confirmed the severity of gyroscope drifts in minute level runs even when the mobile phone is held relatively still to the user =-=[19, 22, 24]-=-. Although there have been many efforts made to address a similar problem (rigid body orientation) in robotics domain [14, 15, 8, 16], our study reveals that directly adopting those approaches cannot ...

Citation Context

...concept of mobile crowdsensing [12] is more and more popular. Recent work has used crowdsensed data to localize users [34] and reduce the calibration efforts of WiFi signatures [23, 38]. Among others =-=[14, 22, 26, 27]-=-, CrowdInside [3] pioneers the efforts of constructing hallway/room shape and connectivity of floor plans. It uses inertial data to build and combine user mobility traces to derive the approximate sha...

Citation Context

...ort walking distance is required [Sen et al. 2013]. Nevertheless, with the emerging trend of crowdsourcingbased indoor localization and navigation [Wu et al. 2013a; Rai et al. 2012; Yang et al. 2012; =-=Shen et al. 2013-=-; Purohit et al. 2013; Jiang et al. 2013], walking traces of diverse users are expected to be calibrated and integrated. Therefore, it would boost the quality of the crowdsourced mobility information ...

Citation Context

...ipped robot. They also advanced earlier work on visual localization based on matching of SIFT descriptors [15] to one using a bag of features that could be stored in mobile phones for quick retrieval [18, 19]. The dataset we introduce in this work is not constrained to robot navigation, as it includes the ego-motion associated with hand-held and wearable devices. Alternative methods: non feature-based and sensor merging. For outdoor navigation, the Global Positioning System (GPS) has been in widespread use for many years. In an indoor context, localization technology is still rapidly evolving [16, 20, 25]. Using visual information is towards the higher end of computational complexity, and possibly the lower-end of reliability; one would certainly seek to support this approach with other forms of sensor such as Received Signal Strength Indication (RSSI) data, magnetometers, and tracking algorithms [16, 18, 19]. In this paper, we seek to explore efficient techniques that could be used to index and compare the visual path information gathered by multiple user journeys, and to measure the potential of vision on its own as a localization mechanism. A biological intuition. Another source of our moti...

Citation Context

... Plan I. INTRODUCTION Location is one of the most important context information in mobile and ubiquitous computing. Recently, wireless indoor localization has attracted extensive research efforts [1]–=-=[11]-=- due to both the need to support indoor location-based services, and the fact that GPS does not work well indoors. However, despite significant progress, an indoor localization system that can be easi...

Citation Context

...e briefly review related work here. Much of the previous work treats the RSSI signals together as a single vector and use some pattern matching technique on these vectors in localization. The work in =-=[5]-=- first stores in database the sequence of Wi-Fi RSS vectors while the user is travelling as patterns along different corridors in an indoor environment. Then the target compares its online measured RS...

Citation Context

...es have recently been proposed that fuse WiFi-based RSS and motion sensor data to simultaneously build a sensor map of the environment and locate the user within this map. e.g. radio fingerprint maps =-=[86, 87]-=-, or organic landmark maps [88]. These approaches focus on location estimation part for an orthogonal problem, where the motion of users cannot be controlled and hence does not involve motion planning...

Citation Context

...ed among peer phones, then maps their locations jointly against WiFi signature map subjecting ranging constraints. Another type of solution somehow utilizes the opportunistic encounters between peers =-=[21, 23]-=-. However, these aforementioned works, more or less, have the following limitations that impede their real-life deployment: 1) Require user movement trace, resulting in that the user cannot obtain his...