bachelorthesis-wifi-sensing-segmentation

Evaluating the Generalisability of Segmentation Methods in Wifi-Sensing

https://github.com/felixdobler/bachelorthesis-wifi-sensing-segmentation

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

✓
CITATION.cff file
Found CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
○
DOI references
○
Academic publication links
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (10.1%) to scientific vocabulary

Keywords

activity-recognition activity-segmentation bachelor-thesis sensing wifi wifi-sensing

Last synced: 6 months ago · JSON representation ·

Repository

Evaluating the Generalisability of Segmentation Methods in Wifi-Sensing

Basic Info

Host: GitHub
Owner: FelixDobler
License: agpl-3.0
Language: Jupyter Notebook
Default Branch: main
Homepage:
Size: 98.2 MB

Statistics

Stars: 1
Watchers: 1
Forks: 0
Open Issues: 0
Releases: 1

Topics

activity-recognition activity-segmentation bachelor-thesis sensing wifi wifi-sensing

Created about 1 year ago · Last pushed 11 months ago

Metadata Files

Readme License Citation

README.md

BachelorThesis - Evaluating the Generalisability of Segmentation Methods in Wifi Sensing

Wi-Fi has become more important in recent years besides the function of providing network connectivity. It is being repurposed for sensing through analysis of Channel State Information. Wi-Fi Sensing can be utilized in many ensing applications like presence and activity detection.
One challenge of Wi-Fi Sensing is to overcome the fact that most approaches still require fixed hardware and are limited to their training environment.
Segementation of CSI is handled as one step towards efficient and accurate sensing.
In this thesis, we test the interoperability of segmentation techniques on data collected with different hardware with a case study. An existing deep learning-based method "DeepSeg" is used to evaluate activity data in a new environment.
With a maximum performance of 90% for segmentation, this approach promises compatibility for different data collection processes.
Further, we highlight challenges in the robustness of existing methods and contribute our tools for public use.

This repository contains the material of my bachelor thesis, including code, dataset, and documentation.

Interactive Activity Segmentation/Labeling Tool

Dateset

The dateset available in the Releases contains 290 labeled activities of 10 different activity kinds, performed by one user. The acquisition methods are detailed in the thesis.

The data can be read using the csidata python package.

Owner

Name: Felix Dobler
Login: FelixDobler
Kind: user
Location: Germany

Repositories: 1
Profile: https://github.com/FelixDobler

Citation (CITATION.cff)

cff-version: 1.2.0
title: >-
  Evaluating the Generalisability of Segmentation Methods in
  Wifi Sensing
message: >-
  If you use this software, please cite it using the
  metadata from this file.
type: software
authors:
  - given-names: Felix
    family-names: Dobler
    orcid: 'https://orcid.org/0009-0006-4846-9352'
    email: felix.dobler.fd@gmail.com
repository-code: >-
  https://github.com/FelixDobler/BachelorThesis-WiFi-Sensing-Segmentation
keywords:
  - WiFi Sensing
  - Segmentation
  - CSI
  - Activity Recognition
license: AGPL-3.0

GitHub Events

Total

Release event: 2
Watch event: 2
Delete event: 1
Push event: 13
Create event: 4

Last Year

Release event: 2
Watch event: 2
Delete event: 1
Push event: 13
Create event: 4

Dependencies

requirements.txt pypi

h5py ==3.11.0
ipykernel ==6.29.5
ipympl ==0.9.4
joblib ==1.4.2
matplotlib ==3.9.0
matplotlib-inline ==0.1.7
numpy ==2.0.1
opencv-python ==4.10.0
pandas ==2.2.2
scipy ==1.14.0
seaborn ==0.13.2

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Open Source Science