Results 11 to 20 of about 4,775 (181)

DS-Mamba: Depthwise separable mamba for hyperspectral image classification.

PLoS ONE
Transformers experience quadratic computational complexity in hyperspectral image (HSI) classification tasks, which can result in error propagation and memory usage issues.
Lin Wei, Huihan Yang, Yuping Yin, Zhiyuan Qu, Haonan Zheng   +4 more
doaj   +3 more sources

Mamba Modulation: On the Length Generalization of Mamba

CoRR
The quadratic complexity of the attention mechanism in Transformer models has motivated the development of alternative architectures with sub-quadratic scaling, such as state-space models. Among these, Mamba has emerged as a leading architecture, achieving state-of-the-art results across a range of language modeling tasks.
Peng Lu 0006, Jerry Huang, Qiuhao Zeng, Xinyu Wang 0061, Boxing Wang, Philippe Langlais, Yufei Cui   +6 more
openaire   +2 more sources

A Survey of Mamba

ACM Transactions on Intelligent Systems and Technology
Deep learning (DL), as a vital technique, has sparked a notable revolution in AI, resulting in a great change in human lifestyles. As one of the most representative DL techniques, the Transformer architecture has empowered numerous advanced models, especially the large language models (LLMs) that comprise billions of parameters ...
Haohao Qu, Liangbo Ning, Rui An, Wenqi Fan, Tyler Derr, Hui Liu, Xin Xu, Qing Li   +7 more
openaire   +2 more sources

Differential Mamba

Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics
Sequence models like Transformers and RNNs often overallocate attention to irrelevant context, leading to noisy intermediate representations. This degrades LLM capabilities by promoting hallucinations, weakening long-range and retrieval abilities, and reducing robustness.
Nadav Schneider, Itamar Zimerman, Eliya Nachmani   +2 more
openaire   +4 more sources

A Probabilistic Fractional Order Physics Informed Mamba Kolmogorov‐Arnold Network for State‐of‐Charge Estimation in Grid‐Connected Battery Energy Storage Systems

Energy Science &Engineering, EarlyView.
This paper introduces the probabilistic fractional‐order Mam‐KAN (PFO‐Mam‐KAN) controller, a physics‐informed gray‐box framework for real‐time battery state‐of‐charge estimation. By unifying efficient Mamba encoders with uncertainty‐aware fractional physics, it achieves superior 0.31% RMSE accuracy and robust grid‐support operation under dynamic ...
Arun Kumar Rawat, Subhash Chandra, Vinay Kumar Deolia   +2 more
wiley   +1 more source

Mamba-Reg: Vision Mamba Also Needs Registers

2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Similar to Vision Transformers, this paper identifies artifacts also present within the feature maps of Vision Mamba. These artifacts, corresponding to high-norm tokens emerging in low-information background areas of images, appear much more severe in Vision Mamba -- they exist prevalently even with the tiny-sized model and activate extensively across ...
Feng Wang 0047, Jiahao Wang 0001, Sucheng Ren, Guoyizhe Wei, Jieru Mei, Wei Shao 0008, Yuyin Zhou, Alan L. Yuille, Cihang Xie   +8 more
openaire   +5 more sources

Attention Based Optimization for 3D Shape Registration

Computer Graphics Forum, EarlyView.
Abstract Transformers are sequence‐to‐sequence architectures originally designed to handle structurally rigid and order‐sensitive data, such as text and images. At their core, they exploit the attention mechanism, which is permutation‐equivariant and relies on computing token‐to‐token relationships.
A. Riva, L. Olearo, S. Melzi
wiley   +1 more source

FaceMamba: Geometry‐Aware Mamba for Efficient Speech‐Driven 3D Facial Animation

Computer Graphics Forum, EarlyView.
Abstract Speech‐driven 3D facial animation plays a pivotal role in immersive digital human applications. Recent works have explored Mamba‐based sequence modeling as an efficient alternative to Transformer, but they often suffer from limited cross‐modal alignment and insufficient control over fine‐grained facial deformations.
Yifan Ge, Zhiqiang Ren, Yazhan Zhang, Haofei Wang, Yawen Zhang, Shengli Zhang   +5 more
wiley   +1 more source

Exploring Prison Safety: Insights From Two Decades of HM Inspectorate of Prisons Surveys in England and Wales

The Howard Journal of Crime and Justice, EarlyView.
ABSTRACT This article presents an analysis of surveys conducted by HM Inspectorate of Prisons for England and Wales of 87,449 adult male prisoners between the years 2000 and 2020. It describes the survey methodology and focuses on the 13,025 people who reported feeling unsafe.
Nicholas Hardwick, Kim Reising, Rosie Meek   +2 more
wiley   +1 more source

Mamba‐Based Deep Learning Model for Automated Periapical Index Classification Using Periapical Radiographs and Clinical Metadata

International Endodontic Journal, EarlyView.
ABSTRACT Aim Apical periodontitis (AP) diagnosis primarily relies on periapical radiographs (PRs) and the Periapical Index (PAI) scoring system. However, existing automated approaches often simplify PAI into binary categories or ignore essential clinical metadata, limiting diagnostic performance and applicability.
Jiyun Lee, Yudam Park, Sunil Kim, Ikbeom Jang   +3 more
wiley   +1 more source