Recognition of Bird Species of Gaoligong Mountain using Transfer Learning Based on VGG-16

The study aims to build a high-quality dataset of local bird images specific to Yunnan province and use it to train and evaluate VGGNet model for bird species recognition. The research will focus on evaluating the performance of advanced VGGNet architecture, in classifying Yunnan local bird species. Firstly, established a dataset named GMBID, which consists of 5 species of bird. Secondly, we will use VGGnet, which had been pre-trained on ImageNet. In this procedure, we will use two ways to train these models: re-training and fine-tuning pre-trained model on training set. In this procedure, we focused on using transfer learning based on VGG16 model. Lastly, the testing of the trained model is performed to examine the performance.