L3DOR: Lifelong 3D Object Recognition
share
3D object recognition has been widely-applied. However, most state-of-the-arts are facing with a fixed recognition task set, which cannot well tackle the new coming data with incremental tasks as human ourselves. Meanwhile, the performance of most state-of-the-art lifelong learning models can be deteriorated easily on previously learned recognition tasks, due to the existing of unordered, large-scale, and irregular 3D geometry data. To address this challenge, in this paper, we propose a Lifelong 3D Object Recognition (i.e., L3DOR framework, which can consecutively learn new 3D object recognition tasks via imitating "human learning". Specifically, the core idea of our proposed L3DOR is to factorize PointNet in a perspective of lifelong learning, while capturing and storing the shared point-knowledge in a perspective of layer-wise tensor factorization architecture. To further transfer the task-specific knowledge from previous tasks to the new coming recognition task, a memory attention mechanism is proposed to connect the current task with relevant previously tasks, which can effectively prevent catastrophic forgetting via soft-transferring previous knowledge. To our best knowledge, this is the first work about using lifelong learning to handle 3D object recognition task without model fine-tuning or retraining. Further, our L3DOR can also be extended to other backbone network (e.g., PointNet++). To the end, comparisons on several point cloud datasets validate that our L3DOR model can reduce averaged 1.68 3.36 times parameters for the overall model, without sacrificing recognition accuracy of each task.
READ FULL TEXT