Cimbiosys: A platform for content-based partial replication
Venugopalan Ramasubramanian; Thomas Rodeheffer; Douglas B. Terry; Meg Walraed-Sullivan; Ted Wobber; Cathy Marshall; Amin Vahdat
People increasingly use multiple devices and Internet services to manage and share information. Since portable devices have limited resources for storage and bandwidth, it is essential to take advantage of proximity and selected replication of content. To this end we present Cimbiosys, a replication platform that permits each device to define its own content-based filtering criteria and to share updates with any other device. Cimbiosys ensures two properties not achieved by previous systems. First, every device stores exactly those items whose latest version meets arbitrary filter criteria that are independent of any hierarchical namespace. Second, every device represents its metadata in a compact form, with state proportional to the number of devices rather than the number of items. Such compact representation enables low synchronization overhead, which permits frequent synchronization even for bandwidth-limited devices. We have implemented Cimbiosys in C# and Mace. We evaluated the performance of the CIMSync protocol in both simulation and using the Mace implementation.
Interactive Techniques for Registering Images to Digital Terrain and Building Models
Billy Chen; Gonzalo Ramos; Eyal Ofek; Michael Cohen; Steve Drucker; david nister
We investigate two interactive techniques for registering an image to 3D digital terrain and building models. Registering an image enables a variety of applications, including slideshows with context, automatic annotation, and photo enhancement. To perform the registration, we investigate two modes of interaction. In the overlay interface, an image is displayed over a 3D view and a usermanually aligns 3D points to points in the image. In the split interface, the image and the 3D view are displayed side-by-side and the user indicates matching points across the two views. Our user study suggests that the overlay interface is more engaging than split, but is less accurate in registration. We then show several applications that make use of the registration data.
An architecture for extensible wireless LANs
Rohan Murty; Jitendra Padhye; Alec Wolman; Matt Welsh
This paper presents Trantor, an architecture for extensible wireless LANs. Trantor enables rapid innovation by removing standardization from the path of introducing new technologies. This is achieved largely by moving the intelligence away from wireless clients and into the infrastructure. In addition to providing extensibility, this approach can also help improve overall network performance through the use of global and historical information. Trantor enables network administrators to impose local policies thereby easing the task of wireless LAN management. In this paper we outline the motivation, vision, and architecture of Trantor.
Online Multi-Label Active Learning for Large-Scale Multimedia Annotation
Xian-Sheng Hua; Guo-Jun Qi
Existing video search engines have not taken the advantages of video content analysis and semantic understanding. Video search in academia uses semantic annotation to approach content-based indexing. We argue this is a promising direction to enable real content-based video search. However, due to the complexity of both video data and semantic concepts, existing techniques on automatic video annotation are still not able to handle large-scale video set and large-scale concept set, in terms of both annotation accuracy and computation cost. To address this problem, in this paper, we propose a scalable framework for annotation-based video search, as well as a novel approach to enable large-scale semantic concept annotation, that is, online multi-label active learning. This framework is scalable to both the video sample dimension and concept label dimension. Large-scale unlabeled video samples are assumed to arrive consecutively in batches with an initial pre-labeled training set, based on which a preliminary multi-label classifier is built. For each arrived batch, a multi-label active learning engine is applied, which automatically selects and manually annotates a set of unlabeled sample-label pairs. And then an online learner updates the original classifier by taking the newly labeled sample-label pairs into consideration. This process repeats until all data are arrived. During the process, new labels, even without any pre-labeled training samples, can be incorporated into the process anytime. Experiments on TRECVID dataset demonstrate the effectiveness and efficiency of the proposed framework.
NetPrints: Diagnosing Home Network Misconfigurations Using Shared Knowledge
Bhavish Aggarwal; Ranjita Bhagwan; Venkata N. Padmanabhan; Geoffrey Voelker
Networks and networked applications depend on several pieces of configuration information to operate correctly. Such information resides in routers, firewalls, and end hosts, among other places. Incorrect information, or misconfiguration, could interfere with the running of networked applications. This problem is particularly acute in consumer settings such as home networks, where there is a huge diversity of network elements and applications coupled with the absence of network administrators. To address this problem, we present NetPrints, a system that leverages shared knowledge in a population of users to diagnose and resolve misconfigurations. Basically, if one user has figured out the fix for a problem, we would like this knowledge made available to another user experiencing the same problem. NetPrints records and aggregates configuration information from a large population of clients, annotates it with compact network problem signatures, looks up the appropriate information when a new client experiences a similar problem, and suggests configuration changes to resolve the problem. NetPrints performs all of these steps automatically, with little human involvement. We evaluate NetPrints in the context of several home networking problems actually reported by users, and find that it is effective in sifting through large volumes of shared configuration data to identify the relevant fix.