Small scale dog face detection using improved Faster RCNN
Lina Mu, Zhijun Shen, Jiangping Liu, Jing Gao · 2022 3rd International Conference on Electronic Communication and Artificial Intelligence (IWECAI) · 2022
In order to achieve fast localization and detection of dog face in intelligent dog management system, a dog face detection algorithm based on improved Faster RCNN was proposed. To obtain the feature extraction backbone network suitable for small size targets, the algorithm extracts the features of the detection target by constructing the residual convolution network, and fuses the features of different layers of the convolution network with the feature pyramid structure. Then, the region proposal network is used to generate and filter the possible regions of the target, and the bilinear interpolation method is used to calculate the bounding box coordinate values to achieve accurate location. Using Tsinghua dog image dataset to train the model, an improved Faster RCNN model adapted to small scale targets is obtained. The experimental results show that the improved method proposed in this paper maintains high precision in the detection of small size dog face, and can realize the detection of different scale objects, which has certain engineering practical value.