Open-world instance-level scene understanding aims to locate and recogni...
Existing 3D scene understanding tasks have achieved high performance on
...
Open-vocabulary scene understanding aims to localize and recognize unsee...
Despite substantial progress in 3D object detection, advanced 3D detecto...
Deep learning approaches achieve prominent success in 3D semantic
segmen...
Large-scale pre-training has been proven to be crucial for various compu...
In this paper, we present a self-training method, named ST3D++, with a
h...
While self-training has advanced semi-supervised semantic segmentation, ...
We present a new domain adaptive self-training pipeline, named ST3D, for...
In this technical report, we present the top-performing LiDAR-only solut...
We focus on Unsupervised Domain Adaptation (UDA) for the task of semanti...
Unsupervised domain adaptation aims to mitigate the domain shift when
tr...