Energy-aware video streaming with QoS control for portable computing devices
NOSSDAV '04 Proceedings of the 14th international workshop on Network and operating systems support for digital audio and video
Autonomous visual model building based on image crawling through internet search engines
Proceedings of the 6th ACM SIGMM international workshop on Multimedia information retrieval
| Hi-index | 0.00 |
In this demo, we show a novel end-to-end video automatic labeling system, which accepts MPEG-1 sequence inputs and generates MPEG-7 XML metadata files. Detections are based on the prior established anchor models. This system has two parts: model training process and labeling process. They are comprised of seven modules: Shot Segmentation, Region Segmentation, Annotation, Feature Extraction, Model Learning, Classification, and XML Rendering.