# fall-detection-2

**Repository Path**: eden33s/fall-detection-2

## Basic Information

- **Project Name**: fall-detection-2
- **Description**: fall detection using machine learning
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-06-10
- **Last Updated**: 2021-06-10

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Fall-Prediction
# 1.fall detection using machine learning (SVM).
# 2.Fall Detection Using CNN (updated):
## Table of Content:
   ### overview
   ### problem statement:
   ### motivation
   ### Technical Aspect
 
## Overview:
   ### This project predicts the fall while walking. Data are manually collected using a mobile app that has an inbuilt accelerometer sensor. Those collected data are used to train the CNN model and finally saved the weights.
## problem statement:
   ### 40% of the traumatic injuries related hospitalization are due to fall, so it is necessary to predict and prevent fall in elderly at early stage. These people always need a care taker and there is a need to improve the independence and quality of life.
## Motivation:
   ### My grandpa became bedridden due to his old age. He may fall from the bed unknowingly. So my mother cared him and protected him from inadvertent fall. Therefore, she is unable to do other activities. From this, I got an idea to develop a mobile app application to detect fall by giving an alert to caretakers when they are away from patients.
## Technical aspect:
   ### Importing necessary libraries.
   
   ### Pre-processing the data:
   #### 1.Balancing the data
   #### 2.Standardizing the data.
   #### 3.Label encoding.
   #### 4.Frame preparation.
   ![](images/frame-preparation.png)
   
   
   ### Training a deep learning model using keras:
   #### 1.2D CNN model
   
   ### Evaluating results:
   #### 1.Confusion matrix
   ![](images/confusion-matrix.png)
   
   #### 2.Model accuracy and model loss graph.