Minding the Gaps in a Video Action Analysis Pipeline
Authors:
- Jia ChenJCJia Chen
- Jiang LiuJLJiang Liu
- Junwei LiangJLJunwei Liang
- Ting-Yao HuTHTing-Yao Hu
- Wei KeWKWei Ke
- Wayner BarriosWBWayner Barrios
- Dong HuangDHDong Huang
- Alexander G. HauptmannAHAlexander G. Hauptmann
- 1 January 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 41-46
Abstract
We present an event detection system, which shares many similarities with standard object detection pipelines. It is composed of four modules: feature extraction, event proposal generation, event classification and event localization. We developed and assessed each module separately by evaluating several candidate options given oracle input using intermediate evaluation metric. This particular process results in a mismatch gap between training and testing when we integrate the module into the complete system pipeline. This results from the fact that each module is trained on clean oracle input, but during testing the module can only receive system generated input, which can be significantly different from the oracle data. Furthermore, we discovered that all the gaps between the different modules can contribute to a decrease in accuracy and they represent the major bottleneck for a system developed in this way. Fortunately, we were able to develop a set of relatively simple fixes in our final system to address and mitigate some of the gaps.Keywords
This publication has 3 references indexed in Scilit:
- Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset2017
- High-Speed Tracking with Kernelized Correlation FiltersIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Improved adaptive Gaussian mixture model for background subtraction2004