Skip to content

alagoz/mrmsnet-tsc

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MRMS-Net / LMRMS-Net

Multi-Scale Representation Networks for Time Series Classification

Official PyTorch implementation of MRMS-Net and LMRMS-Net, two convolutional architectures designed for efficient and robust time series classification (TSC).

The models exploit multi-scale temporal representations using parallel convolutional branches while maintaining strong computational efficiency.

This repository provides:

  • Reference implementations of MRMS-Net and LMRMS-Net
  • A minimal training pipeline
  • A one-command demo using a sample dataset
  • Architecture diagrams

The full experimental framework used in the paper (large-scale evaluation across 142 datasets) is not included in this repository.


Overview

Time series classification often requires capturing patterns occurring at multiple temporal scales. MRMS-Net and LMRMS-Net address this by combining parallel convolutional filters with different receptive fields.

MRMS-Net

A multi-scale architecture with three convolutional branches capturing short-, medium-, and long-term temporal patterns.

Key characteristics:

  • Parallel convolutions: k = 3, 5, 7
  • Feature fusion block
  • Global average pooling
  • Lightweight yet expressive representation learning

LMRMS-Net

A computationally efficient variant designed for fast inference.

Key characteristics:

  • Two multi-scale branches
  • Optional early-exit inference
  • Reduced parameter count
  • Designed for low-latency scenarios

Architecture

Architecture

Left: MRMS-Net architecture with three parallel convolutional branches. Right: LMRMS-Net lightweight architecture with optional early exit.


Installation

Clone the repository:

git clone https://github.com/alagoz/mrmsnet-tsc
cd mrmsnet-tsc

Install dependencies:

pip install -r requirements.txt

Quick Demo (One Command)

Run the demo training script:

python demo/run_demo.py

This script will:

  1. Load a small example dataset
  2. Initialize MRMS-Net
  3. Train the model for a few epochs
  4. Print training progress

The demo runs in under a minute on CPU.


Repository Structure

MRMS-Net-lsnet-tsc
│
├── models
│   ├── msnet.py
│   └── lsnet.py
│
├── demo
│   └── run_demo.py
│
├── data
│   └── ucr_loader.py
│
├── figures
│   └── architecture.png
│
├── requirements.txt
└── README.md

Models

Model Description
MRMS-Net Multi-scale CNN with 3 convolution branches
LMRMS-Net Lightweight multi-scale CNN with optional early exit

Example Usage

from models.msnet import MSNet

model = MSNet(
    in_channels=1,
    n_classes=5
)

Citation

If you use MRMS-Net or LMRMS-Net in your research, please cite:

@article{alagoz2026msnet,
  title={Multi-Scale Representation Networks for Time Series Classification},
  author={Alagoz, Celal},
  journal={},
  year={2026}
}

License

MIT License


Contact

For questions regarding the models or implementation, please open a GitHub issue.

About

MSNet / LS-Net: Multi-Scale Representation Networks for Time Series Classification Reference PyTorch implementation of MSNet and LS-Net, two lightweight multi-scale CNN architectures for time series classification. The repository provides model implementations and runnable demos on UCR datasets.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages