A Normalized Cross-Correlation Convolutional Neural Network (CNN-NCC) for Exemplar-Based Object Detection
Yaya Wihardi, Winda Mauli Kristy, Erlangga, Arjon Turnip, Intan Nurma Yulita, Endroyono

Department of Computer Science, Universitas Pendidikan Indonesia
Artificial Intelligence and Big Data Research Center, Universitas Padjadjaran
Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember

Abstract

On this study we try to build an exemplar-based object detector based on Convolutional Neural Network with Normalized Cross Correlation matching. Our proposed method is did not need to re-train for identified unknown object that have not been acknowledge by model during training stage. The result show that the model achieves 0.449 mAP on COCO dataset.

Keywords: CNN, Normalized Cross Correlation, Object Detection, Exemplar Based

Topic: Computer Science