We present LongLoRA, an efficient fine-tuning approach that extends the
...
False negatives (FN) in 3D object detection, e.g., missing predictions
o...
Category-level 6D pose estimation aims to predict the poses and sizes of...
LiDAR-based 3D point cloud recognition has benefited various application...
3D object detectors usually rely on hand-crafted proxies, e.g., anchors ...
3D scenes are dominated by a large number of background points, which is...
Recent advances in 2D CNNs and vision transformers (ViTs) reveal that la...
In this work, we present a conceptually simple yet effective framework f...
Non-uniformed 3D sparse data, e.g., point clouds or voxels in different
...
Separating 3D point clouds into individual instances is an important tas...
Differentiable architecture search (DARTS) marks a milestone in Neural
A...
In this paper, we present a conceptually simple, strong, and efficient
f...
We propose Scale-aware AutoAug to learn data augmentation policies for o...
In this report, we present our object detection/instance segmentation sy...
Object detectors commonly vary quality according to scales, where the
pe...
Recently, numerous handcrafted and searched networks have been applied f...
A single-point feature has shown its effectiveness in object detection.
...
Object detectors are usually equipped with networks designed for image
c...
Neural architecture search (NAS) methods have been proposed to release h...
Designing neural architectures is a fundamental step in deep learning
ap...
Neural architecture search (NAS) is an important task in network design,...