DeepKalPose: An Enhanced Deep-Learning Kalman Filter for Temporally Consistent Monocular Vehicle Pose Estimation
This work addresses the challenge of accurate and consistent vehicle pose estimation from monocular video, which is crucial for applications like autonomous driving, but it appears incremental as it builds upon existing deep-learning and Kalman filter methods.
The paper tackled the problem of temporally consistent monocular vehicle pose estimation in video by introducing DeepKalPose, which integrates a bi-directional Kalman filter with a learnable motion model, resulting in improved pose accuracy and robustness, particularly for occluded or distant vehicles, as validated on the KITTI dataset.
This paper presents DeepKalPose, a novel approach for enhancing temporal consistency in monocular vehicle pose estimation applied on video through a deep-learning-based Kalman Filter. By integrating a Bi-directional Kalman filter strategy utilizing forward and backward time-series processing, combined with a learnable motion model to represent complex motion patterns, our method significantly improves pose accuracy and robustness across various conditions, particularly for occluded or distant vehicles. Experimental validation on the KITTI dataset confirms that DeepKalPose outperforms existing methods in both pose accuracy and temporal consistency.