Beomjoo Seo

Efficient disk replacement and data migration algorithms for large disk subsystems

Random data placement, which is efficient and scalable for large-scale storage systems, has recently emerged as an alternative to traditional data striping. In this report, we study the disk replacement problem (DRP) to find a sequence of disk additions and removals for a storage system, while migrating the data and respecting the following constraints: (1) the data is initially balanced across the existing distributed disk configuration, (2) the data must again be balanced across the new configuration, and (3) the data migration cost must be minimized. In practice, migrating data from old disks to new devices is complicated by the fact that the total number of disks connected to the storage system is often limited by a fixed number of available slots and not all the old and new disks can be connected at the same time. This article presents solutions for both cases where the number of disk slots is either unconstrained or constrained.

Dynamic scheduling on video transcoding for MPEG DASH in the cloud environment

The Dynamic Adaptive Streaming over HTTP (referred as MPEG DASH) standard is designed to provide high quality of media content over the Internet delivered from conventional HTTP web servers. The visual content, divided into a sequence of segments, is made available at a number of different bitrates so that an MPEG DASH client can automatically select the next segment to download and play back based on current network conditions. The task of transcoding media content to different qualities and bitrates is computationally expensive, especially in the context of large-scale video hosting systems. Therefore, it is preferably executed in a powerful cloud environment, rather than on the source computer (which may be a mobile device with limited memory, CPU speed and battery life). In order to support the live distribution of media events and to provide a satisfactory user experience, the overall processing delay of videos should be kept to a minimum. In this paper, we propose a novel dynamic scheduling methodology on video transcoding for MPEG DASH in a cloud environment, which can be adapted to different applications. The designed scheduler monitors the workload on each processor in the cloud environment and selects the fastest processors to run high-priority jobs. It also adjusts the video transcoding mode (VTM) according to the system load. Experimental results show that the proposed scheduler performs well in terms of the video completion time, system load balance, and video playback smoothness.

An experimental study of video uploading from mobile devices with HTTP streaming

Mobile video traffic is growing rapidly in networks due to the continuing user adoption of smartphones and tablet computers. While video viewing is now prevalent on such devices, they also easily enable the recording and uploading of videos for quick publishing on popular video sharing websites. Dynamic Adaptive Streaming over HTTP, or DASH, is a media streaming standard that has recently been developed by the Motion Pictures Experts Group (MPEG) and which has gained attention for its ability to enable media players to render videos with high quality under various network conditions. MPEG-DASH has been ratified at the end of 2011 and is now also known as ISO/IEC 23009--1. It is noteworthy that the focus of the initial standard is limited to the efficient server-to-client distribution of videos. In our study we examine the common challenges that manifest themselves during the client-to-server uploading of mobile videos, for example such issues as unstable wireless connections and delayed video availability. We propose a new approach that provides compatibility with DASH and at the same time improves content availability by reducing the end-to-end delay from the recording time of mobile videos to the publishing of the first segment of the multi-bitrate encoded versions through a careful pipelining of the overall process. Our approach features (1) the use of segmentation of videos on the mobile device before uploading and (2) segment-wise transcoding and transformatting on the server side. Therefore, our solution does not require any dedicated encoder for live events while achieving semi-realtime live streaming and providing multi-bitrate content for user-generated videos from smartphones. We report on the performance of our prototype system which uses Android and iOS client devices.

Multi-video summary and skim generation of sensor-rich videos in geo-space

User-generated videos have become increasingly popular in recent years. Due to advances in camera technology it is now very easy and convenient to record videos with mobile devices, such as smartphones. Here we consider an application where users collect and share a large set of videos that are related to a geographic area, say a city. Such a repository can be a great source of information for prospective tourists when they plan to visit a city and would like to get a preview of its main areas. The challenge that we address is how to automatically create a preview video summary from a large set of source videos. The main features of our technique are that it is fully automatic and leverages meta-data sensor information which is acquired in conjunction with videos. The meta-data is collected from GPS and compass sensors and is used to describe the viewable scenes of the videos. Our method then proceeds in three steps through the analysis of the sensor data. First, we generate a single video summary. Shot boundaries are detected based on different motion types of camera movements and key frames are extracted related to motion patterns. Second, we build video skims for popular places (i.e., hotspots) aiming to provide maximal coverage of hotspot areas with minimal redundancy (per-spot multi-video summary). Finally, the individual hotspot skims are linked together to generate a pleasant video tour that visits all the popular places (multi-spot multi-video summary).

Keyframe presentation for browsing of user-generated videos on map interfaces

To present user-generated videos that relate to geographic areas for easy access and browsing it is often natural to use maps as interfaces. A common approach is to place thumbnail images of video keyframes in appropriate locations. Here we consider the challenge of determining which keyframes to select and where to place them on the map. Our proposed technique leverages sensor-collected meta-data which are automatically acquired as a continuous stream together with the video. Our approach is able to detect interesting regions and objects (hotspots) and their distances from the camera in a fully automated way. Meaningful keyframes are adaptively selected based on the popularity of the hotspots. Our experiments show very promising results and demonstrate excellent utility for the users.

Degrees of Freedom of Millimeter Wave Full-Duplex Systems With Partial CSIT

The degrees of freedom (DoF) of L-path poor scattering full-duplex (FD) systems is studied in which a FD base station having M transmit antennas and M receive antennas supports a set of half-duplex mobile stations (MSs). Assuming no self-interference, a hybrid scheduling is proposed achieving the optimal sum DoF with partial channel state information at the transmitter side as the number of MSs increases. In particular, the proposed scheme combines a zero-forcing beamforming for uplink and a random transmit beamforming for downlink with opportunistic scheduling. It is shown that the optimal sum DoF of 2M is asymptotically achievable as long as the number of MSs scales faster than snrmin(M-1,L)+M, where snr denotes the signal-to-noise ratio.

Content vs. Context: Visual and Geographic Information Use in Video Landmark Retrieval

Due to the ubiquity of sensor-equipped smartphones, it has become increasingly feasible for users to capture videos together with associated geographic metadata, for example the location and the orientation of the camera. Such contextual information creates new opportunities for the organization and retrieval of geo-referenced videos. In this study we explore the task of landmark retrieval through the analysis of two types of state-of-the-art techniques, namely media-content-based and geocontext-based retrievals. For the content-based method, we choose the Spatial Pyramid Matching (SPM) approach combined with two advanced coding methods: Sparse Coding (SC) and Locality-Constrained Linear Coding (LLC). For the geo-based method, we present the Geo Landmark Visibility Determination (GeoLVD) approach which computes the visibility of a landmark based on intersections of a cameras field-of-view (FOV) and the landmarks geometric information available from Geographic Information Systems (GIS) and services. We first compare the retrieval results of the two methods, and discuss the strengths and weaknesses of each approach in terms of precision, recall and execution time. Next we analyze the factors that affect the effectiveness for the content-based and the geo-based methods, respectively. Finally we propose a hybrid retrieval method based on the integration of the visual (content) and geographic (context) information, which is shown to achieve significant improvements in our experiments. We believe that the results and observations in this work will enlighten the design of future geo-referenced video retrieval systems, improve our understanding of selecting the most appropriate visual features for indexing and searching, and help in selecting between the most suitable methods for retrieval based on different conditions.

Sensor-rich video exploration on a map interface

Result presentations from searches into video repositories is still a challenging problem. Current systems usually display a ranked list that shows the first frame of each video. Users then explore the videos one-by-one. In our recent work we have investigated the fusion of captured video with a continuous stream of sensor meta-data. These so-called sensor-rich videos can conveniently be captured with todays smartphones. Importantly, the recorded sensor-data streams enable processing and result resentation in novel and useful ways. In this demonstration we present a system that provides an integrated solution to present videos based on keyframe extraction and interactive, map-based browsing. As a key feature, the system automatically computes popular places based on the collective information from all the available videos. For each video it then extracts keyframes and renders them at their proper location on the map synchronously with the video playback. All the processing is performed in real-time, which allows for an interactive exploration of all the videos in a geographic area.

Orientation data correction with georeferenced mobile videos

Similar to positioning data, camera orientation information has become a powerful contextual feature utilized by a number of GIS and social media applications. Such auxiliary information facilitates higher-level semantic analysis and management of video assets in such applications, e.g., video summarization and video indexing systems. However, it is problematic that raw sensor data collected from current mobile devices is often not accurate enough for subsequent geospatial analysis. To date, an effective orientation data correction system for mobile video content has been lacking. Here we present a content-based approach that improves the accuracy of noisy orientation sensor measurements generated by mobile devices in conjunction with video acquisition. Our preliminary experimental results demonstrate significant accuracy enhancements which benefit upstream sensor-aided GIS applications to access video content more precisely.

Automatic positioning data correction for sensor-annotated mobile videos

Video associated positioning data has become a useful contextual feature to facilitate analysis and management of media assets in GIS and social media applications. Moreover, with todays sensor-equipped mobile devices, the location of a camera can be continuously acquired in conjunction with the captured video stream without much difficulty. However, most sensor information collected from mobile devices is not highly accurate due to two main reasons: (a) the varying surrounding environmental conditions during data acquisition, and (b) the use of low-cost, consumer-grade sensors in current mobile devices. In this paper, we enhance the noisy positioning data generated by smartphones during video recording by analyzing typical error patterns for real collected data and introducing two robust algorithms, based on Kalman filtering and weighted linear least square regression, respectively. Our experimental results demonstrate significant benefits of our methods, which help upstream sensor-aided applications to access media content precisely.