MambaNeXt-YOLO: A Hybrid State Space Model for Real-time Object Detection

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lei, Xiaochun, Wu, Siqi, Wu, Weilin, Jiang, Zetao
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911073665286144
author Lei, Xiaochun
Wu, Siqi
Wu, Weilin
Jiang, Zetao
author_facet Lei, Xiaochun
Wu, Siqi
Wu, Weilin
Jiang, Zetao
contents Real-time object detection is a fundamental but challenging task in computer vision, particularly when computational resources are limited. Although YOLO-series models have set strong benchmarks by balancing speed and accuracy, the increasing need for richer global context modeling has led to the use of Transformer-based architectures. Nevertheless, Transformers have high computational complexity because of their self-attention mechanism, which limits their practicality for real-time and edge deployments. To overcome these challenges, recent developments in linear state space models, such as Mamba, provide a promising alternative by enabling efficient sequence modeling with linear complexity. Building on this insight, we propose MambaNeXt-YOLO, a novel object detection framework that balances accuracy and efficiency through three key contributions: (1) MambaNeXt Block: a hybrid design that integrates CNNs with Mamba to effectively capture both local features and long-range dependencies; (2) Multi-branch Asymmetric Fusion Pyramid Network (MAFPN): an enhanced feature pyramid architecture that improves multi-scale object detection across various object sizes; and (3) Edge-focused Efficiency: our method achieved 66.6% mAP at 31.9 FPS on the PASCAL VOC dataset without any pre-training and supports deployment on edge devices such as the NVIDIA Jetson Xavier NX and Orin NX.
format Preprint
id arxiv_https___arxiv_org_abs_2506_03654
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MambaNeXt-YOLO: A Hybrid State Space Model for Real-time Object Detection
Lei, Xiaochun
Wu, Siqi
Wu, Weilin
Jiang, Zetao
Computer Vision and Pattern Recognition
Artificial Intelligence
Real-time object detection is a fundamental but challenging task in computer vision, particularly when computational resources are limited. Although YOLO-series models have set strong benchmarks by balancing speed and accuracy, the increasing need for richer global context modeling has led to the use of Transformer-based architectures. Nevertheless, Transformers have high computational complexity because of their self-attention mechanism, which limits their practicality for real-time and edge deployments. To overcome these challenges, recent developments in linear state space models, such as Mamba, provide a promising alternative by enabling efficient sequence modeling with linear complexity. Building on this insight, we propose MambaNeXt-YOLO, a novel object detection framework that balances accuracy and efficiency through three key contributions: (1) MambaNeXt Block: a hybrid design that integrates CNNs with Mamba to effectively capture both local features and long-range dependencies; (2) Multi-branch Asymmetric Fusion Pyramid Network (MAFPN): an enhanced feature pyramid architecture that improves multi-scale object detection across various object sizes; and (3) Edge-focused Efficiency: our method achieved 66.6% mAP at 31.9 FPS on the PASCAL VOC dataset without any pre-training and supports deployment on edge devices such as the NVIDIA Jetson Xavier NX and Orin NX.
title MambaNeXt-YOLO: A Hybrid State Space Model for Real-time Object Detection
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2506.03654