Video Clustering via Multidimensional Time-Series Analysis
| Autor: | Stolbovyi Mykhailo, Mashtalir Sergii, Kobylin Oleg |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Big data processing
Parsing Computer science business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Video sequence Pattern recognition 010103 numerical & computational mathematics 02 engineering and technology computer.software_genre 01 natural sciences Fault detection and isolation 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence 0101 mathematics Equivalence (formal languages) Time series business Cluster analysis computer Video retrieval |
| Zdroj: | ICIME |
| DOI: | 10.1145/3149572.3149599 |
| Popis: | Video analysis is one of the most complex scientific areas related to the big data processing. This is due to the non-structuredness of video data. That makes it difficult to understand the inner frames relationships. Video sequences parsing with reference to frames semantic discrepancy search is one of the main challenges in content based video retrieval. Search for such matches allows to split the video into equivalence or tolerance classes that contain semantically similar frames. Approaches to video clustering on the base of a multidimensional time series analysis are proposed. Suggested techniques are concerned with the clusters boundaries (video shots) fault detection. Video clustering performance depending on feature selections is considered. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|