MambaTron: Efficient Cross-Modal Point Cloud Enhancement Using Aggregate Selective State Space Modeling
Year: 2025
Type: conference-paper
Abstract: Point cloud enhancement is the process of generating a high-quality point cloud from an incomplete input. This is done by filling in the missing details from a reference like the ground truth via regression, for example. In addition to unimodal image and point cloud reconstruction, we focus on the task of view-guided point cloud completion, where we gather the missing information from an image, which represents a view of the point cloud and use it to generate the output point cloud. With the recent research efforts surrounding state-space models, originally in natural language processing and now in 2D and 3D vision, Mamba has shown promising results as an efficient alternative to the self-attention mechanism. However, there is limited re-search towards employing Mamba for cross-attention be-tween the image and the input point cloud, which is crucial in multi-modal problems. In this paper, we introduce Mamba Tron, a Mamba- Transformer cell that serves as a building block for our network which is capable of uni-modal and cross-modal reconstruction which includes view-guided point cloud completion. We explore the benefits of Mamba's long-sequence efficiency coupled with the Trans-former's excellent analytical capabilities through MambaT-ron. This approach is one of the first attempts to implement a Mamba-based analogue of cross-attention, especially in computer vision. Our model demonstrates a degree of performance comparable to the current state-of-the-art techniques while using a fraction of the computation resources.
Authors: Sai Tarun Inaganti, Gennady Petrenko
Institution: University of Minnesota
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4409917052