Search papers, labs, and topics across Lattice.
This paper addresses the challenge of long-tailed class distributions in point cloud dataset distillation, which has been overlooked by existing methods that primarily focus on geometric and representation aspects. The authors introduce a novel framework that incorporates Adaptive Synthetic Budgeting and 3D Long-Tailed Distribution Matching to effectively allocate synthetic sample budgets and optimize point clouds for both global class distributions and intra-class diversity. Experimental results show a significant improvement in classification accuracy, achieving a 7.0-point increase on ShapeNet55 compared to state-of-the-art techniques.
Long-tailed point cloud dataset distillation can boost classification accuracy by 7 points, challenging the limits of existing methods.
Dataset distillation compresses large-scale datasets into compact synthetic sets while preserving their training utility, enabling efficient 3D point cloud training. Current point cloud dataset distillation methods only tackle geometric and representation challenges while ignoring the distributional imbalance prevalent in point cloud datasets where both training and test splits follow long-tailed class distributions. To our knowledge, we present the first study on long-tailed point cloud dataset distillation. Rather than focusing primarily on geometric and representation properties or simply constructing a class-balanced synthetic set, our framework explicitly accounts for long-tailed class distributions via two core modules. First, we design Adaptive Synthetic Budgeting to allocate class-wise synthetic budgets according to class quantity and the expected benefit of additional synthetic samples. Given the allocated budgets, we further design 3D Long-Tailed Distribution Matching to optimize synthetic point clouds through Global-Local Feature Alignment and Prior-Aware Supervision. The former preserves both global class distributions and diverse intra-class structures, while the latter provides class-dependent expert supervision to keep tail-class samples recognizable while maintaining diverse head-class patterns. Extensive experiments demonstrate the effectiveness of our method, lifting classification accuracy by 7.0 points on ShapeNet55 against state-of-the-art methods.
