🎓 A Deep Learning Based Human Activity Recognition Using Wi-Fi Signals

Major Project :

This project uses a Raspberry Pi and Wi-Fi router to collect Channel State Information (CSI) data, which tracks changes in Wi-Fi signals caused by human movements, to train a model that recognizes different human activities.

Project Workflow

• Hardware: Integrating Hardware components for data collection.

• Data Collection: Collecting CSI data using Raspberry Pi(Rx) and Router(Tx).

• Preprocessing: Filtering and transforming the CSI data.

• Visualization: Visualizing the CSI data before and after preprocessing.

• Feature Extraction and Annotation: Extracting features from the preprocessed data and annotating the data for training.

• Model Training: Building model architecture and training the model to recognize different human activities.

• UI: Developing an application integrating all the components.

  graph TD
      A[Hardware Setup] --> B[CSI Data Collection]
      B --> C[Preprocessing]
      C --> D[Visualization]
      D --> E[Feature Extraction and Annotation]
      E --> F[Model Training]
      F --> G[UI]
  

  Loading

---

Repository Structure

• Implementation: Source Code. under development!!

• Reference Papers: Researches Highly Relevant to our work.