IMUWiFine: End-to-End Sequential Indoor Localization Dataset

This dataset accompanies the research paper "End-to-End Sequential Indoor Localization Using Smartphone Inertial Sensors and WiFi". It provides data for training and evaluating an end-to-end indoor localization system.

Dataset Description

The IMUWiFine dataset contains data collected from smartphone inertial measurement units (IMUs) and WiFi received signal strength indicators (RSSIs). The data is organized into training, validation, and testing sets, enabling robust model evaluation. The dataset is designed for use with the provided source code, which implements a sequential indoor localization architecture based on a stack of ReLU, LSTM, and regression layers. The architecture takes IMU and WiFi RSSI readings as input and outputs estimated (x, y, z) position coordinates. The dataset can be downloaded from ISSAI. The dataset is divided into train, test, and validation folders.

Dataset Structure

The dataset consists of three folders: train, val, and test. Each folder contains the data required for training, validation, and testing respectively. The exact format of the data within these folders is not specified here, but details can be found in the associated source code.

Usage

The provided source code utilizes the PyTorch framework. The paths to the training, validation, and testing data folders must be specified within the train.py script. The experiment name should also be specified in train.py before training.

Citation

While a citation is provided in the original content, it is omitted here as requested.