ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection

Xu, Zhenbo; Zhang, Wei; Ye, Xiaoqing; Tan, Xiao; Yang, Wei; Wen, Shifeng; Ding, Errui; Meng, Ajin; Huang, Liusheng

doi:10.1609/aaai.v34i07.6945

Cited by 73 publications

(33 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1) 2D-based methods [7,54,30,39,57,49,38] first detect 2D bounding box proposals and then regress instancewise 3D boxes. Stereo-RCNN [30] extended Faster R-CNN [43] for stereo-inputs to associate left and right images.…”

Section: Related Workmentioning

confidence: 99%

“…Stereo-RCNN [30] extended Faster R-CNN [43] for stereo-inputs to associate left and right images. Disp R-CNN [49] and ZoomNet [57] incorporated extra instance segmentation mask and part location map to improve detection quality. However, the final performance is limited by the recall of 2D detection algorithms, and 3D geometry information is not fully utilized.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector

Guo¹,

Shi²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Stereo-based 3D detection aims at detecting 3D objects from stereo images, which provides a low-cost solution for 3D perception. However, its performance is still inferior compared with LiDAR-based detection algorithms. To detect and localize accurate 3D bounding boxes, LiDARbased detectors encode high-level representations from Li-DAR point clouds, such as accurate object boundaries and surface normal directions. In contrast, high-level features learned by stereo-based detectors are easily affected by the erroneous depth estimation due to the limitation of stereo matching. To solve the problem, we propose LIGA-Stereo (LiDAR Geometry Aware Stereo Detector) to learn stereo-based 3D detectors under the guidance of high-level geometry-aware representations of LiDAR-based detection models. In addition, we found existing voxel-based stereo detectors failed to learn semantic features effectively from indirect 3D supervisions. We attach an auxiliary 2D detection head to provide direct 2D semantic supervisions. Experiment results show that the above two strategies improved the geometric and semantic representation capabilities. Compared with the state-of-the-art stereo detector, our method has improved the 3D detection performance of cars, pedestrians, cyclists by 10.44%, 5.69%, 5.97% mAP respectively on the official KITTI benchmark. The gap between stereo-based and LiDAR-based 3D detectors is further narrowed. The code is available at https: //xy-guo.github.io/liga/.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector

Guo¹,

Shi²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…To draw and visualize a object's orientation from a bird's eye view perspective, the object's α will need to be converted into r y . With the object's x and z location information, we can convert r y to α or vice versa [40,41,42], using the formula:…”

Section: Alpha (Local/allocentric Rotation)mentioning

confidence: 99%

SoK: Vehicle Orientation Representations for Deep Rotation Estimation

Tu¹,

Peng²,

Leung³

et al. 2021

Preprint

View full text Add to dashboard Cite

In recent years, there is an influx of deep learning models for 3D vehicle object detection. However, little attention was paid to orientation prediction. Existing research work proposed various vehicle orientation representation methods for deep learning, however a holistic, systematic review has not been conducted. Through our experiments, we categorize and compare the accuracy performance of various existing orientation representations using the KITTI 3D object detection dataset [1], and propose a new form of orientation representation: Tricosine. Among these, the 2D Cartesian-based representation [cos(θ), sin(θ)], or Single Bin, achieves the highest accuracy, with additional channeled inputs (positional encoding and depth map) not boosting prediction performance. Our code is published on GitHub: https://github.com/umd-fire-coml/KITTI-orientationlearning

show abstract

“…A robust autonomous driving system requires its LiDAR-based detector to reliably handle different environmental conditions, e.g., geographic locations and weather conditions. While 3D detection has received increasing interest in recent years, most existing works [79,7,10,11,16,23,25,26,27,29,30,36,41,42,49,50,51,53,63,64,65,67,68,62,78] The regions of "missing points" are irregular in shape.…”

Section: Introductionmentioning

confidence: 99%