Not available / text

Not available / text

Unreliable networks can severely hamper transmission of video data. In applications requiring minimal latency, video frames must be compressed using intraframe techniques. We develop a video codec suitable for robot teleoperation over unreliable networks with high packet loss rates. The codec combines a foveated image compression algorithm, Embedded Foveation Image enCoding (EFIC), with a Forward Error Correction (FEC) code tuned to network performance. Foveation, or spatially variant image resolution, allows very high compression levels while preserving the most important image characteristics. By tightly integrating an FEC within the codec we are able to virtually eliminate dropped frames independent of the network protocol. We find that the new codec supports much higher video quality than another intraframe compression technique, Motion JPEG (M-JPEG). / text

published_or_final_version / abstract / toc / Electrical and Electronic Engineering / Master / Master of Philosophy

published_or_final_version / Electrical Engineering / Master / Master of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

The Best MPhil Thesis in the Faculties of Architecture, Arts, Business& Economics, Education, Law and Social Sciences (University of HongKong), Li Ka Shing Prize, 1997-1999 / published_or_final_version / Music / Master / Master of Philosophy

The current study examined the effectiveness of a video modeling procedure on a basicmath skill acquisition in students with Autism Spectrum Disorder (ASD) using a multipleprobe across students design. Participants were four kindergarten/first grade students in aself-contained classroom in an urban public school. All met the criteria for ASD prior tobeginning the study. The independent variable included a video model teaching a basicmath skill. The dependent variable was correct student response. Results indicated thatone of the four students responded to the intervention immediately and these skillsmaintained over time. However, the other three students did not respond to videomodeling instruction. Conclusions regarding the effectiveness of the video modelingprocedure for teaching math skills were inconclusive. Implications for further researchare addressed.

Imitation is a core deficit often observed in children diagnosed with autism. Video modeling has been shown to be effective for teaching children with autism a variety of skills, but there is little research demonstrating the effectiveness of this technique with core skills such as imitation. The purpose of this study was to examine the effectiveness of a video modeling intervention to teach a preschool-age child with autism to imitate novel and acquired actions (with and without objects) in natural contexts (i.e., songs and toy play activities). A general case approach was used to examine the instructional universe of common preschool songs in order to select the exemplars that were most likely to facilitate generalization. In addition to video modeling, additive components that included highlighting the critical features of the video examples and prompting/fading were required to demonstrate a functional relationship. Experimental control was evident in a multiple baseline design across three imitation activities. The results are discussed with reference to previous research, future research directions, and implications for practice in educational settings.

With the emergence of peer-to-peer video-on-demand systems, new avenues for keeping track of and subsequently meeting user needs and desires have arisen. Based on the idea of contextual priming, we introduce a new frame-work, SUE, that takes advantage of the intimate level of user profiling afforded by the internet as well as the linear and segmented nature of p2p technology to determine a user's exact on-screen experience at any given time interval. This allows us to more accurately determine the type of information a user is likely to be more receptive to. Our design differs from other existing systems in two ways: (a) the level of granularity it can support, accommodating factors from both the user's on-screen and physical environment in making its recommendations and (b) in addressing some of the shortcomings seen in current applications, such as those imposed by coarse user profiling and faulty associations. In order to examine the viability of our framework, we provide a high level design specifying its incorporation into an existing p2p video system, the BitVampire project.