| Idea Name: |
InFALLible |
| Slogan: |
Automated and Adaptive System |
| Supervisor Name: |
Dr. Ahsan Shahzad |
| Supervisor Designation: |
Assistant Professor |
| Supervisor School: |
College of Electrical and Mechanical Engineering |
| Supervisor Department: |
Computer and Software Department |
| Contact number: |
03185055986 |
| Email ID: |
[email protected] |
| Abstract: |
| We aim to develop an automated and adaptive fall detection with an alert message system along with cloud based deep learning deployment for personalization and edge AI for local decisions. |
| What is the unmet need in society that your idea will fulfill ? |
| Any individual can experience a fall, maybe from a cardiac arrest or a neurological disorder. Sometimes these falls can be hazardous causing unconsciousness for a long period of time. These falls can be detected automatically by an abnormal movement pattern and the caregivers can be alerted . |
| Who needs it ? How many would benefit ? |
| More than 37.3 million people around the globe experience unwanted falls each year. Out of them, more than 6.4 million people succumb to the jaws of death. More than 5.1 million people who experience falls are from low and middle income countries. In Pakistan, there are 3000 such patients annually. |
| How will the solution works |
| Smartphone internal accelerometer sensor will be continuously used for acquiring the data for fall detection. Critical or false alarm data will be saved on our local database. User feedback will be taken to assign some labels. After enough accumulation of data, it will be sent to the cloud via the internet for retraining and calculation of personalized hyperparameters. After classification of a fall, an alert message will be sent to the caregivers along with the exact location of the patient. |
| Who are your competitors ? How is your solution different |
| There are no local/tested verified applications and existing solutions are unreliable. Hence, no such potential competitors exist in the market. Our application will be reliable and execute with efficient performance. Highly personalized application with user friendly GUI available internationally. |
| Status: |
new |
| Entry Date & Time: |
2020-12-31 (1756) |
