Vision mambas have demonstrated strong performance with linear complexity to the number of vision tokens. Their efficiency results from processing image tokens sequentially. However, most existing methods employ patch-based image tokenization and then flatten them into 1D sequences for causal processing, which ignore the intrinsic 2D structural correlations of images. It is also difficult to extract global information by sequential processing of local patches. In this paper, we propose a global image serialization method to transform the image into a sequence of causal tokens, which contain global information of the 2D image. We first convert the image from the spatial domain to the frequency domain using Discrete Cosine Transform (DCT) and then arrange the pixels with corresponding frequency ranges. We further transform each set within the same frequency band back to the spatial domain to obtain a series of images before tokenization. We construct a vision mamba model, GlobalMamba, with a causal input format based on the proposed global image serialization, which can better exploit the causal relations among image sequences. Extensive experiments demonstrate the effectiveness of our GlobalMamba, including image classification on ImageNet-1K, object detection on COCO, and semantic segmentation on ADE20K.
Voxel Mamba: Group-Free State Space Models for Point Cloud based 3D Object Detection
The Hong Kong Polytechnic University(香港理工大学);Center for Artificial Intelligence and Robotics, HKISI, CAS(中国科学院香港智能科学研究所人工智能与机器人中心);Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)
BiSegMamba: Efficient Bidirectional Tri-Oriented Mamba for 3D Medical Image Segmentation
School of Computer Science and Engineering, Beihang University(北航计算机科学与工程学院);State Key Laboratory of Virtual Reality Technology and Systems, Beihang University(北航虚拟现实技术与系统国家重点实验室)
TrackingMiM: Efficient Mamba-in-Mamba Serialization for Real-time UAV Object Tracking
Department of Electronic and Electrical Engineering, Southern University of Science and Technology(电子与电气工程系,南方科技大学);Peng Cheng Laboratory(鹏城实验室);Department of Applied Mathematics and Theoretical Physics, University of Cambridge(应用数学与理论物理系,剑桥大学);State Grid Corporation of China(国家电网公司);East China Institute of Computing Technology(东部计算技术研究所);Shenzhen Key Laboratory of Robotics and Computer Vision(深圳机器人与计算机视觉重点实验室)
DEPFusion: Dual-Domain Enhancement and Priority-Guided Mamba Fusion for UAV Multispectral Object Detection
School of Information Engineering, Guangdong University of Technology(广东工业大学信息工程学院);School of Electronic and Information Engineering, South China University of Technology(华南理工大学电子信息学院);School of Artificial Intelligence, XiDian University(西安电子科技大学人工智能学院)
Mantis: Mamba-native Tuning is Efficient for 3D Point Cloud Foundation Models
Xi’an Jiaotong University(西安交通大学);School of Artificial Intelligence and Robotics, Hunan University(湖南大学人工智能与机器人学院);University of Western Australia(西澳大学)
SpectralMamba-UNet: Frequency-Disentangled State Space Modeling for Texture-Structure Consistent Medical Image Segmentation
College of Computer Science(计算机科学学院);Sichuan Normal University(四川师范大学);Zhejiang University(浙江大学);Ant Group(蚂蚁集团);State Key Laboratory of Public Big Data(公共大数据国家重点实验室);Guizhou University(贵州大学)
3DTMDet: A Dual-Path Synergy Network of Transformer and SSM for 3D Object Detection in Point Clouds
School of Electronic and Optical Engineering(电子与光学工程学院);The 28th Research Institute of China Electronics Technology Group Corporation(中国电子科技集团第二十八研究所);College of Astronautics(航天学院);School of Information and Communication Engineering(信息与通信工程学院);State key Laboratory of Extreme Environment Optoelectronic Dynamic Measurement Technology and Instrument(极端环境光电动态测量技术与仪器国家重点实验室)
FaRMamba: Frequency-based learning and Reconstruction aided Mamba for Medical Segmentation
School of Artificial Intelligence and Computer Science, Nantong University, China(人工智能与计算机科学学院,南通大学);School of Information Science and Technology, Nantong University, China(信息科学与技术学院,南通大学)