Automated segmentation of video files with structurally complex content
Shih-Fu Chang, Ph.D.
Enables shot-independent segmentationLess susceptible to noise, such as camera movement, than shot-based technologiesAble to perform segmentation that is independent of canonical eventsPatent information:Patent Issued (US 6,865,226 Tech Ventures Reference: IR M02-070
Automated augmentation of sports or instructional videos with indices that can be used to facilitate browsingEliminating extraneous content in video filesIdentification of event episodes of potential interest in long surveillance videosAnalysis of genetically modified model organism behavior videos in neuroscience researchClassification of different sequences of a video into several desired categories
Shih-Fu Chang, Ph.D.
USA
