Multiple object tracking (MOT) has become a very important task in the field of rail transit. Recently, the most effective method in MOT tasks is tracking by detection. Because there are many pedestrians under a single camera in the railway station, the occluded pedestrians will generate detection boxes with low-score. Once these low-score bounding boxes are mistakenly classified as background by the detector, the traditional tracking model may discard these boxes, resulting in poor tracking effect. In order to solve this problem effectively, this paper proposes a novel model with Generic Association Mechanism (GAM), which can pay attention to every detection box instead of only focusing on the highscore ones. This method utilizes the similarity between these low-score detection boxes and trajectories to restore true objects and filter out the background detection boxes. More attention can be focused on these low-score detection boxes through this method, which contribute the most to the tracking effect. This method can efficiently avoid the loss of true objects and fragmented trajectories of occluded objects. Finally, we test this proposed model with other models on several public datasets, and the results show that the proposed method achieves more significant performance improvement.
Model iteration with new data is important to improve generalization of the model. In general, there are two methods to deal with model incremental update: (a) retraining the model with merging all data together and (b) training a separate model with the new data based on transfer learning. However, the above methods are either time-consuming or suffering from over-fitting problems when the sample size of new data is small. To address this practical issue, we propose a new iteration model, the IterationNet, which can learn features of new data while maintain the performance on the old data. It is a new model iteration method based on knowledge distillation which adds consistency network and truncate L1 regularization. In classifying fake avatar images of Weibo users, IterationNet extremely decreased training time from 8 hours to 5 minutes while the accuracy rate is only reduced from 96% to 91% comparing to training with merged data. Compared with transfer learning, IterationNet showed increased accuracy rate by 21 percent with similar training time.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.