A Survey on Mamba Architecture for Vision Applications
It addresses scalability challenges in computer vision for researchers and practitioners, but is incremental as it reviews existing advancements.
This paper surveys the Mamba architecture, which uses state-space models to tackle the quadratic complexity of Transformers in vision tasks, achieving linear scalability and improved contextual awareness for applications like object detection and video understanding.
Transformers have become foundational for visual tasks such as object detection, semantic segmentation, and video understanding, but their quadratic complexity in attention mechanisms presents scalability challenges. To address these limitations, the Mamba architecture utilizes state-space models (SSMs) for linear scalability, efficient processing, and improved contextual awareness. This paper investigates Mamba architecture for visual domain applications and its recent advancements, including Vision Mamba (ViM) and VideoMamba, which introduce bidirectional scanning, selective scanning mechanisms, and spatiotemporal processing to enhance image and video understanding. Architectural innovations like position embeddings, cross-scan modules, and hierarchical designs further optimize the Mamba framework for global and local feature extraction. These advancements position Mamba as a promising architecture in computer vision research and applications.