ABSTRACT
This work presents a data-driven adaptive windowing approach to accelerate video content extraction in DNN-based Complex Event Processing (CEP) systems. The CEP windows continuously monitor low-level content of incoming video frames and exploit interframe correlations to accelerate the overall DNN content extraction process. The two main contributions are: 1) technique to create micro-batches of similar frames within the window by measuring dissimilarities among them, and 2) optimal frame resolution within micro-batches under specified accuracy thresholds for fast model processing. The initial experimental results show that our adaptive micro-batching approach improves 3.75X model throughput execution while maintaining application-level latency bounds under required accuracy constraints.
- Bifet, A. and Gavaldà, R. 2007. Learning from Time-Changing Data with Adaptive Windowing. SDM (2007).Google Scholar
- Cugola, G. and Margara, A. 2012. Processing flows of information. ACM Computing Surveys. 44, 3 (2012), 1--62.Google ScholarDigital Library
- He, K., Zhang, X., Ren, S. and Sun, J. 2016. Deep Residual Learning for Image Recognition. IEEE CVPR (2016).Google Scholar
- Image Pyramids -- OpenCV Documentation: https://bit.ly/2mjqkdp.Google Scholar
- Kang, D., Emmons, J., Abuzaid, F., Bailis, P. and Zaharia, M. 2017. NoScope: Optimizing Neural Network Queries over Video at Scale. VLDB (2017).Google Scholar
- Yadav, P. and Curry, E. 2019. VEKG: Video Event Knowledge Graph to Represent Video Streams for Complex Pattern Matching. IEEE Graph Computing (2019).Google Scholar
- Yadav, P. and Curry, E. 2019. VidCEP: Complex Event Processing Framework to Detect Spatiotemporal Patterns in Video Streams. IEEE BigData (2019).Google Scholar
Index Terms
- Data-Driven Windows to Accelerate Video Stream Content Extraction in Complex Event Processing
Recommendations
High-performance complex event processing framework to detect event patterns over video streams
Middleware '19: Proceedings of the 20th International Middleware Conference Doctoral SymposiumComplex Event Processing (CEP) is an event processing paradigm capable of detecting patterns over streaming data in real-time. Presently, CEP systems have key challenges to preform matching over video streams due to their unstructured data model and ...
Stream reasoning and complex event processing in ETALIS
On linked spatiotemporal data and geo-ontologiesAddressing dynamics and notifications in the Semantic Web realm has recently become an important area of research. Run time data is continuously generated by multiple social networks, sensor networks, various on-line services and so forth. How to get ...
Dual-Paradigm Stream Processing
ICPP '18: Proceedings of the 47th International Conference on Parallel ProcessingExisting stream processing frameworks operate either under data stream paradigm processing data record by record to favor low latency, or under operation stream paradigm processing data in micro-batches to desire high throughput. For complex and mutable ...
Comments