Awesome

MambaTSR: You Only Need 90k Parameters for Traffic Sign Recognition

Our work “MambaTSR: You Only Need 90k Parameters for Traffic Sign Recognition” has been accepted by Neurocomputing(JCR Q1) 🎉🎉🎉 and the paper will online soon!

[Paper]

Zhihao Chen, Yiyuan Ge, Mingxin Yu, Qing Yue, Lianqing Zhu

⭐If this work is helpful for you, please help star this repo. Thanks!🤗

Getting Started

Installation

Step 1: Clone the VMamba repository:

To get started, first clone the MambaTSR repository and navigate to the project directory:

git clone https://github.com/1024AILab/MambaTSR.git
cd MambaTSR

Step 2: Environment Setup:

MambaTSR recommends setting up a conda environment and installing dependencies via pip. Use the following commands to set up your environment:

Create and activate a new conda environment

conda create -n vmamba
conda activate vmamba

Install Dependencies

pip install -r requirements.txt
cd kernels/selective_scan && pip install .

Samples

AdditiveGaussianNoise

mean = 0

<img src="images/samples/AdditiveGaussianNoise_0/1.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_0/2.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_0/3.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_0/4.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_0/5.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_0/6.jpg" height="250"/>

mean = 120

<img src="images/samples/AdditiveGaussianNoise_120/1.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_120/2.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_120/3.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_120/4.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_120/5.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_120/6.jpg" height="250"/>

mean = -120

<img src="images/samples/AdditiveGaussianNoise_d120/1.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_d120/2.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_d120/3.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_d120/4.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_d120/5.jpg" height="250"/><img src="images/samples/AdditiveGaussianNoise_d120/6.jpg" height="250"/>

Rain

s = 3.0

<img src="images/samples/weather_rain_3/1.jpg" height="250"/><img src="images/samples/weather_rain_3/2.jpg" height="250"/><img src="images/samples/weather_rain_3/3.jpg" height="250"/><img src="images/samples/weather_rain_3/4.jpg" height="250"/><img src="images/samples/weather_rain_3/5.jpg" height="250"/><img src="images/samples/weather_rain_3/6.jpg" height="250"/>

s = 7.0

<img src="images/samples/weather_rain_7/1.jpg" height="250"/><img src="images/samples/weather_rain_7/2.jpg" height="250"/><img src="images/samples/weather_rain_7/3.jpg" height="250"/><img src="images/samples/weather_rain_7/4.jpg" height="250"/><img src="images/samples/weather_rain_7/5.jpg" height="250"/><img src="images/samples/weather_rain_7/6.jpg" height="250"/>

You can download the datasets at 链接：https://pan.baidu.com/s/1T8sp81USl5AClQmvf0IzPw

提取码：gdr8