An efficient hand detection method based on convolutional neural network

In this paper, we propose an efficient method for detecting human hands based on the architecture of YOLO [1]. By utilizing the spatial-transfer connection (STC) between high-level layers and low-level layers, the multi-scale features from different layers can be aggregated for detecting the hands. Experimental results on the Hand Dataset [2] show that the proposed architecture can improve efficiency and achieves better results than YOLO when detecting small objects such as human hands.