Geometry and uncertainty-aware 3d point cloud class-incremental semantic segmentation

Y Yang, M Hayat, Z **, C Ren… - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Despite the significant recent progress made on 3D point cloud semantic segmentation, the
current methods require training data for all classes at once, and are not suitable for real-life …

Continual learning on 3D point clouds with random compressed rehearsal

M Zamorski, M Stypułkowski, K Karanowski… - Computer Vision and …, 2023 - Elsevier
Contemporary deep neural networks offer state-of-the-art results when applied to visual
reasoning, eg, in the context of 3D point cloud data. Point clouds are an important data type …

Cross-Domain Incremental Feature Learning for ALS Point Cloud Semantic Segmentation with Few Samples

M Dai, S **ng, Q Xu, P Li, J Pan… - IEEE Transactions on …, 2024 - ieeexplore.ieee.org
Feature learning of airborne laser scanning (ALS) point clouds is challenged by both the
limited annotated samples and imbalanced class distribution. An intuitive way involves …

Rethinking task-incremental learning baselines

MS Hossain, P Saha, TF Chowdhury… - 2022 26th …, 2022 - ieeexplore.ieee.org
It is common to have continuous streams of new data that need to be introduced in the
system in real-world applications. The model needs to learn newly added capabilities (future …

A Benchmark Grocery Dataset of Realworld Point Clouds From Single View

SV Sheshappanavar, T Anvekar… - … Conference on 3D …, 2024 - ieeexplore.ieee.org
Fine-grained grocery object recognition is an important computer vision problem with broad
applications in automatic checkout, in-store robotic navigation, and assistive technologies …

ReFu: Recursive Fusion for Exemplar-Free 3D Class-Incremental Learning

Y Yang, L Zhong, H Zhuang - arxiv preprint arxiv:2409.12326, 2024 - arxiv.org
We introduce a novel Recursive Fusion model, dubbed ReFu, designed to integrate point
clouds and meshes for exemplar-free 3D Class-Incremental Learning, where the model …

Continual Learning in 3D Point Clouds: Employing Spectral Techniques for Exemplar Selection

H Resani, B Nasihatkon, MA Jazi - arxiv preprint arxiv:2409.08388, 2024 - arxiv.org
We introduce a novel framework for Continual Learning in 3D object classification (CL3D).
Our approach is based on the selection of prototypes from each class using spectral …