Abstract:
To improve urban transportation systems, transport mode identification (TMI) must infer user trajectories’ travel modes. Existing TMI methods use mobility features from densely sampled GPS trajectory points (1 second per GPS point) or IMU sensor data (accelerometer, gyroscope, rotation vector) to improve accuracy. These drain mobile device batteries, though. This paper introduces the Multi-Scale Attributes Attention (MSAA) model, a deep learning framework that extracts discriminating trajectory features from GPS data without increasing its sampling rate. The proposed model partitions trajectories into scales and extracts local attribute latent representations at each scale. The MSAA model uses Convolutional Neural Network (CNN) to capture spatial correlation of trajectory segments and attention mechanism to select the best local attributes on different trajectory scales to characterize the various transport modes. Since the learned latent local attributes are significantly different from the global features (e.g. average/min/max travel speeds, which are measurable quantities), an ensemble model based on Neural Decision Forest (NDF) is used to fuse the heterogeneous features of measurable and non-measurable elements to determine the transport mode. The proposed approach outperforms several state-of-the-art baselines in accuracy by 0.76% to 6.4% in real-world datasets. Multi-scale local attributes also complement global features. Local attributes improved detection performance by 2.3% compared to global features.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
