Performance Improvement on Object Detection for the Specific Domain Object Detecting
Currently, various studies are underway to improve the performance of Object Detection technology. In the proposed system, rather than classifying all fields as a whole, we want to check the performance of the domain by performing additional learning on the specific domain. In this paper, we propose the performance improvement in the object detection system for detecting human in the coastal image that shot by drone.
KeywordsObject detection Single Shot MultiBox Detector Human detection Computer vision Image classification
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF2017R1D1A1B03030033).
- 2.Redmon, J., Farhadi, A.: YOLO9000: better, faster, stronger. arXiv:1612.08242 (2016)
- 3.Choi, H., Kim, S., Bae, G., Noh, Y., Lee, D.: Real-time object recognition and reporting system. In: Korea Software Congress 2017, pp. 2107–2109 (2017)Google Scholar
- 4.Lee, J., Kim, D., Kim, J.: CPU-based YOLO performance enhancement technique for CCTV real-time object detection. In: Korea Computer Congress 2018, pp. 870–872 (2018)Google Scholar