Real-time tele-immersion requires low latency, synchronized multi-camera capture. Prior high definition (HD) capture systems were bulky. We in vestigate the suitability of using flocks of smartphone cameras for tele-immersion. Smartphones can potentially integrate HD capture and streaming into a single portable package. However, they are designed for archiving the captured video into a movie. Hence, we create a sequence of H.264 movies and stream them. We lower the capture delay by reducing the number of frames in each movie segment. Increasing the number of movie segments adds compression overhead. Smartphone video encoders do not sacrifice video quality to lower the compression latency or the stream size. On an iPhone 4S, our application that uses published APIs streams 1920x1080 videos at 16.5 fps with a delay of 712 msec between a real-life event and displaying an uncompressed bitmap of this event on a local laptop. For comparison, the bulky Cisco Tandberg required 300 msec delay. Stereoscopic video from two unsynchronized smartphones showed minimal visual artifacts in an indoor teleconference setting.

Video tends to be imbalanced as a medium. Typically, content creators invest enormous effort creating work that is then watched passively. However, learning tasks require that users not only consume video but also engage, interact with, and repurpose content. Furthermore, to promote learning across domains where content creators are not necessarily videographers, it is important that capture tools facilitate creation of interactive content. In this paper, we describe some early experiments toward this goal. Specifically, we describe a needfinding study involving interviews with amateur video creators as well as our experience with an early prototype to support expository capture and access. Our findings led to a system redesign that can incorporate a broad set of video-creation and interaction styles.

The expressiveness of touch input can be increased by detecting additional finger pose information at the point of touch such as finger rotation and tilt. PointPose is a prototype that performs finger pose estimation at the location of touch using a short-range depth sensor viewing the touch screen of a mobile device. We present an algorithm that extracts finger rotation and tilt from a point cloud generated by a depth sensor oriented towards the device's touchscreen. The results of two user studies we conducted show that finger pose information can be extracted reliably using our proposed method. We show this for controlling rotation and tilt axes separately and also for combined input tasks using both axes. With the exception of the depth sensor, which is mounted directly on the mobile device, our approach does not require complex external tracking hardware, and, furthermore, external computation is unnecessary as the finger pose extraction algorithm can run directly on the mobile device. This makes PointPose ideal for prototyping and developing novel mobile user interfaces that use finger pose estimation.

A panel at ACM Multimedia 2012 addressed research successes in the past 20 years. While the panel focused on the past, this article discusses successes since the ACM SIGMM 2003 Retreat and suggests research directions in the next ten years. While significant progress has been made, more research is required to allow multimedia to impact our everyday computing environment. The importance of hardware changes on future research directions is discussed. We believe ubiquitous computing—meaning abundant computation and network bandwidth—should be applied in novel ways to solve multimedia grand challenges and continue the IT revolution of the past century.

Unlike text, copying and pasting parts of video documents is challenging. Yet, the huge amount of video documents now available in the form of how-to tutorials begs for simpler techniques that allow users to easily copy and paste fragments of video materials into new documents. We describe new direct video manipulation techniques that allow users to quickly copy and paste content from video documents such as how-to tutorials into a new document. While the video plays, users interact with the video canvas to select text regions, scrollable regions, slide sequences built up across many frames, or semantically meaningful regions such as dialog boxes. Instead of relying on the timeline to accurately select sub-parts of the video document, users navigate using familiar selection techniques such as mouse-wheel to scroll back and forward over a video region where content scrolls, double-clicks over rectangular regions to select them, or clicks and drags over textual regions of the video canvas to select them. We describe the video processing techniques that run in real-time in modern web browsers using HTML5 and JavaScript; and show how they help users quickly copy and paste video fragments into new documents, allowing them to efficiently reuse video documents for authoring or note-taking.

Capturing book images is more convenient with a mobile phone camera than with more specialized flat-bed scanners or 3D capture devices. We built an application for the iPhone 4S that captures a sequence of hi-res (8 MP) images of a page spread as the user sweeps the device across the book. To do the 3D dewarping, we implemented two algorithms: optical flow (OF) and structure from motion (SfM). Making further use of the image sequence, we examined the potential of multi-frame OCR. Preliminary evaluation on a small set of data shows that OF and SfM had comparable OCR performance for both single-frame and multi-frame techniques, and that multi-frame was substantially better than single-frame. The computation time was much less for OF than for SfM.

People often use more than one query when searching for information; they also revisit search results to re-find information. These tasks are not well-supported by search interfaces and web browsers. We designed and built a Chrome browser extension that helps people manage their ongoing information seeking. The extension combines document and process metadata into an interactive representation of the retrieved documents that can be used for sense-making, for navigation, and for re-finding documents.

Exploratory search is a complex, iterative information seeking activity that involves running multiple queries, finding and examining many documents. We introduced a query preview interface that visualizes the distribution of newly-retrieved and re-retrieved documents prior to showing the detailed query results. When evaluating the preview control with a control condition, we found effects on both people’s information seeking behavior and improved retrieval performance. People spent more time formulating a query and were more likely to explore search results more deeply, retrieved a more diverse set of documents, and found more different relevant documents when using the preview. With more time spent on query formulation, higher quality queries were produced and as consequence the retrieval results improved; both average residual precision and recall was higher with the query preview present.

Existing user interfaces for the configuration of large shared displays with multiple inputs and outputs usually do not allow users easy and direct configuration of the display's properties such as window arrangement or scaling. To address this problem, we are exploring a gesture-based technique for manipulating display windows on shared display systems. To aid target selection under noisy tracking conditions, we propose VoroPoint, a modified Voronoi tessellation approach that increases the selectable target area of the display windows. By maximizing the available target area, users can select and interact with display windows with greater ease and precision.

Collaboration technologies must support information sharing between collaborators, but must also take care not to share too much information or share information too widely. Systems that share information without requiring an explicit action by a user to initiate the sharing must be particularly cautious in this respect. Presence systems are an emerging class of applications that support collaboration. Through the use of pervasive sensors, these systems estimate user location, activities, and available communication channels. Because such presence data are sensitive, to achieve wide-spread adoption, sharing models must reflect the privacy and sharing preferences of their users. This paper looks at the role that privacy-preserving aggregation can play in addressing certain user sharing and privacy concerns with respect to presence data. We define conditions to achieve CollaPSE (Collaboration Presence Sharing Encryption) security, in which (i) an individual has full access to her own data, (ii) a third party performs computation on the data without learning anything about the data values, and (iii) people with special privileges called “analysts” can learn statistical information about groups of individuals, but nothing about the individual values contributing to the statistic other than what can be deduced from the statistic. More specifically, analysts can decrypt aggregates without being able to decrypt the individual values contributing to the aggregate. Based in part on studies we carried out that illustrate the need for the conditions encapsulated by CollaPSE security, we designed and implemented a family of CollaPSE protocols. We analyze their security, discuss efficiency tradeoffs, describe extensions, and review more recent privacy-preserving aggregation work.

Although longer queries can produce better results for information seeking tasks, people tend to type short queries. We created an interface designed to encourage people to type longer queries, and evaluated it in two Mechanical Turk experiments. Results suggest that our interface manipulation may be effective for eliciting longer queries.

People frequently capture photos with their smartphones, and some are starting to capture images of documents. However, the quality of captured document images is often lower than expected, even when applications that perform post-processing to improve the image are used. To improve the quality of captured images before post-processing, we developed a Smart Document Capture (SmartDCap) application that provides real-time feedback to users about the likely quality of a captured image. The quality measures capture the sharpness and framing of a page or regions on a page, such as a set of one or more columns, a part of a column, a figure, or a table. Using our approach, while users adjust the camera position, the application automatically determines when to take a picture of a document to produce a good quality result. We performed a subjective evaluation comparing SmartDCap and the Android Ice Cream Sandwich (ICS) camera application; we also used raters to evaluate the quality of the captured images. Our results indicate that users find SmartDCap to be as easy to use as the standard ICS camera application. Additionally, images captured using SmartDCap are sharper and better framed on average than images using the ICS camera application.

Motivated by the addition of gyroscopes to a large number of new smart phones, we study the effects of combining accelerometer and gyroscope data on the recognition rate of motion gesture recognizers with dimensionality constraints. Using a large data set of motion gestures we analyze results for the following algorithms: Protractor3D, Dynamic Time Warping (DTW) and Regularized Logistic Regression (LR). We chose to study these algorithms because they are relatively easy to implement, thus well suited for rapid prototyping or early deployment during prototyping stages. For use in our analysis, we contribute a method to extend Protractor3D to work with the 6D data obtained by combining accelerometer and gyroscope data. Our results show that combining accelerometer and gyroscope data is beneficial also for algorithms with dimensionality constraints and improves the gesture recognition rate on our data set by up to 4%.

We describe direct video manipulation interactions applied to screen-based tutorials. In addition to using the video timeline, users of our system can quickly navigate into the video by mouse-wheel, double click over a rectangular region to zoom in and out, or drag a box over the video canvas to select text and scrub the video until the end of a text line even if not shown in the current frame. We describe the video processing techniques developed to implement these direct video manipulation techniques, and show how there are implemented to run in most modern web browsers using HTML5's CANVAS and Javascript.

Video is becoming a prevalent medium for e-learning. Lecture videos contain useful information in both the visual and aural channels: the presentation slides and lecturer's speech respectively. To extract the visual information, we apply video content analysis to detect slides and optical character recognition (OCR) to obtain their text. Automatic speech recognition (ASR) is used similarly to extract spoken text from the recorded audio. These two text sources have distinct characteristics and relative strengths for video retrieval. We perform controlled experiments with manually created ground truth for both the slide and spoken text from more than 60 hours of lecture video. We compare the automatically extracted slide and spoken text in terms of accuracy relative to ground truth, overlap with one another, and utility for video retrieval. Experiments reveal that automatically recovered slide text and spoken text contain different content with varying error profiles. Additional experiments demonstrate higher precision video retrieval using automatically extracted slide text.