Point-SRA: Self-Representation Alignment for 3D Representation Learning

TL;DR

Point-SRA introduces a novel 3D representation learning method that aligns multi-level features through self-distillation and probabilistic modeling, improving performance across various 3D tasks.

Contribution

It proposes a dual self-representation alignment mechanism at both MAE and MeanFlow Transformer levels, addressing limitations of fixed mask ratios and point-wise reconstruction.

Findings

Outperforms Point-MAE by 5.37% on ScanObjectNN.

Achieves 96.07% mean IoU on intracranial aneurysm segmentation.

Surpasses MaskPoint by 5.12% AP@50 in 3D object detection.

Abstract

Masked autoencoders (MAE) have become a dominant paradigm in 3D representation learning, setting new performance benchmarks across various downstream tasks. Existing methods with fixed mask ratio neglect multi-level representational correlations and intrinsic geometric structures, while relying on point-wise reconstruction assumptions that conflict with the diversity of point cloud. To address these issues, we propose a 3D representation learning method, termed Point-SRA, which aligns representations through self-distillation and probabilistic modeling. Specifically, we assign different masking ratios to the MAE to capture complementary geometric and semantic information, while the MeanFlow Transformer (MFT) leverages cross-modal conditional embeddings to enable diverse probabilistic reconstruction. Our analysis further reveals that representations at different time steps in MFT also exhibit complementarity. Therefore, a Dual Self-Representation Alignment mechanism is proposed at both the MAE and MFT levels. Finally, we design a Flow-Conditioned Fine-Tuning Architecture to fully exploit the point cloud distribution learned via MeanFlow. Point-SRA outperforms Point-MAE by 5.37% on ScanObjectNN. On intracranial aneurysm segmentation, it reaches 96.07% mean IoU for arteries and 86.87% for aneurysms. For 3D object detection, Point-SRA achieves 47.3% AP@50, surpassing MaskPoint by 5.12%.