High Dimensional Frustum PointNet for 3D Object Detection from Camera, LiDAR, and Radar
2020, Wang, Leichen, Chen, Tianbai, Anklam, Carsten, Goldlücke, Bastian
Fusing the raw data from different automotive sensors for real-world environment perception is still challenging due to their different representations and data formats. In this work, we propose a novel method termed High Dimensional Frustum PointNet for 3D object detection in the context of autonomous driving. Motivated by the goals data diversity and lossless processing of the data, our deep learning approach directly and jointly uses the raw data from the camera, LiDAR, and radar. In more detail, given 2D region proposals and classification from camera images, a high dimensional convolution operator captures local features from a point cloud enhanced with color and temporal information. Radars are used as adaptive plug-in sensors to refine object detection performance. As shown by an extensive evaluation on the nuScenes 3D detection benchmark, our network outperforms most of the previous methods.
Sparse-PointNet : See Further in Autonomous Vehicles
2021, Wang, Leichen, Goldlücke, Bastian
Since the density of LiDAR points reduces significantly with increasing distance, popular 3D detectors tend to learn spatial features from dense points and ignore very sparse points in the far range. As a result, their performance degrades dramatically beyond 50 meters. Motivated by the above problem, we introduce a novel approach to jointly detect objects from multimodal sensor data, with two main contributions. First, we leverage PointPainting  to develop a new key point sampling algorithm, which encodes the complex scene into a few representative points with approximately similar point density. Further, we fuse a dynamic continuous occupancy heatmap to refine the final proposal. In addition, we feed radar points into the network, which allows it to take into account additional cues. We evaluate our method on the widely used nuScenes dataset. Our method outperforms all state-of-the-art methods in the far range by a large margin and also achieves comparable performance in the near range.
Radar Ghost Target Detection via Multimodal Transformers
2021, Wang, Leichen, Giebenhain, Simon, Anklam, Carsten, Goldlücke, Bastian
Ghost targets caused by inter-reflections are by design unavoidable in radar measurements, and it is challenging to distinguish these artifact detections from real ones. In this letter, we propose a novel approach to detect radar ghost targets by using LiDAR data as a reference. For this, we adopt a multimodal transformer network to learn interactions between points. We employ self-attention to exchange information between radar points, and local crossmodal attention to infuse information from surrounding LiDAR points. The key idea is that a ghost target should have higher semantic affinity with the reflected real target than the other ones. Extensive experiments on nuScenes  show that our method outperforms the baseline method on radar ghost target detection by a large margin.