Efficient 3D Object Detection from Point Cloud with Noise Conditioned Score Network
Efficient 3D object detection using noise conditioned score network for accurate votes and object proposals.
Reviving Dense BEV Frameworks for 3D Object Detection: BEVNeXt
Enhancing dense BEV frameworks for accurate 3D object detection with BEVNeXt.
LiDAR Object Detection의 Scaling Adversarial Robustness인 ScAR
3D 객체 감지에서의 Scaling Adversarial Robustness의 중요성
CN-RMA: Combined Network with Ray Marching Aggregation for 3D Indoor Object Detection from Multi-view Images
CN-RMA introduces an innovative approach for 3D indoor object detection from multi-view images, achieving state-of-the-art performance by combining reconstruction and detection networks with occlusion-aware feature aggregation.
BEVENet: A Convolutional Approach to Efficient 3D Object Detection in Bird's-Eye-View Space for Autonomous Driving
BEVENet, a novel convolutional-only architecture, achieves state-of-the-art efficiency in 3D object detection for autonomous driving by leveraging Bird's-Eye-View (BEV) space and outperforms computationally intensive Vision Transformer (ViT)-based methods.
Progressive Multi-Modal Fusion for Robust 3D Object Detection (ProFusion3D)
ProFusion3D is a novel LiDAR-camera fusion framework for robust 3D object detection that leverages a progressive fusion strategy across different views (BEV and PV) and levels (intermediate features and object queries), enhanced by a self-supervised pre-training method for improved data efficiency.
ROA-BEV: Enhancing 3D Object Detection in Bird's-Eye View Using 2D Region-Oriented Attention
ROA-BEV improves the accuracy of vision-based 3D object detection in autonomous driving by using 2D region-oriented attention to help the network focus on areas where objects are likely to exist, thereby reducing interference from background information.
CVCP-Fusion: An Investigation into Implicit Depth Estimation for 3D Bounding Box Prediction Using Cross-View Transformers and CenterPoint
Combining LiDAR and camera data for 3D object detection using implicit depth estimation from cross-view transformers, while theoretically promising, struggles to achieve accurate height predictions in 3D bounding boxes, suggesting the need for explicit depth calculation methods.
© 2024 by Linnk AI