Identification tracking network training method, tracking method, system, device and medium

• Main Authors: YAN KEFEI, TANG HUIQIONG, ZHANG XIAOGANG, HE HANBING, YOU XUEQIANG, ZHENG JIANCHAO, WANG YUYANG, QIU MIAOLIN, YANG ZONGHAN
• Format: Patent
• Language: Chinese; English
• Published: 24.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a recognition tracking network training method, a recognition tracking method, a recognition tracking system, a recognition tracking device and a medium. The method comprises the following steps: acquiring a current frame image, a previous frame image of the current frame and a next frame image of the current frame of a visual sensor; inputting the target frame image into a target detection network to obtain a target frame image corresponding to the target frame image; performing first processing on the current frame image and the historical frame image to obtain images of the current frame image and the historical frame image; inputting the current frame image and the historical frame image into a twin neural network to obtain a first feature matrix and a second feature matrix; obtaining a fusion matrix; inputting the fusion matrix into a reinforcement learning model to obtain a prediction image of a frame after the current frame; determining a loss function according to the predicted