Real-time Object Detection with FPGA Using CenterNet
| Autor: | Dmitry Telpukhov, Roman A. Solovyev, I. I. Romanova, Ilya A. Mkrtchan, Alexander G. Kustov |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
SIMPLE (military communications protocol)
Artificial neural network business.industry Computer science 02 engineering and technology 010501 environmental sciences Object (computer science) 01 natural sciences Object detection 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Layer (object-oriented design) business Fixed-point arithmetic Field-programmable gate array Encoder Computer hardware 0105 earth and related environmental sciences |
| Zdroj: | 2021 IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering (ElConRus). |
| DOI: | 10.1109/elconrus51938.2021.9396702 |
| Popis: | the paper proposes methodology for transferring architecture of modern neural network CenterNet to FPGA. CenterNet is a OneStage object detector that is used to detect and locate objects in images. Although this neural network has simple decoder, it shows good performance in terms of accuracy. Very high operation speed of the neural network hardware is achieved due to the choice of suitable encoder and efficient hardware implementation of both the decoder and the last layer with object filtering, and also transition to fixed-point arithmetic. At this, quality of the obtained predictions remains high. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|