Steam School 2018
Because of lack of infrastructure and social problems, children get to go through lot of security and safety issues in schools and outside. We ,here, propose an Auditory diary Saathi, which is Artificial intelligence based listening ear to children and detects the mood variations of them on day to day basis. Also, we intend to correlate the color choices of children with their voice data so that we can generate data to validate our studies to contribute to existing research.For prototype, the device has a microphone connected to a Raspberry PI, which is programmed to convert speech to text and identify some keywords.Once these keywords are found in the child's story, a third party (counsellor or NGO) is alerted.
Pictures, key skills and team roles
Vivek Martin,Manas Chhabra, Ivan D Lepcha
This project is about ensuring safety of children in schools by promising them a platform to express themselves .
Process and motivation
Existing safety measures are basically in the firm of CCTV camera or security personals both of which are mostly post-incident help and in most cases CCTV footages are not directly taken as a foolproof evidence by the court.Our solution has the potential tackle the problem before the child gets into a possible danger.
It is when one of the team member highlighted that sexual harassment has been a very acute problem among children and there are not enough counsellors for every child. What solutions could be devised so the problem can be reported and victims can get educated ? Three of us are school teachers. One of us faced similar issue when one of his student faced an unfortunate incident. The child showed the traits of confusion and anxiety. Also, loss of attention was also observed. The basic challenges are how to categorize the different levels of abuse so that there is no mis-reporting of the incident. Second challenge is that kids feel safe enough to report any mis-happening, that everyone is educated not to repeat such abuse.
Human Activity Recognition example using TensorFlow on smartphone sensors dataset and an LSTM RNN (Deep Learning algo). Classifying the type of movement amongst six activity categories - Guillaume ...
guillaume-chevalier/LSTM-Human-Activity-Recognition • github.com
User persona, data collection
Body Atlas Reveals Where We Feel Happiness and Shame - D-brief
New research shows emotional states are associated with distinct bodily sensations regardless of culture and so specific that they can be mapped.
Body Atlas Reveals Where We Feel Happiness and Shame - D-brief • blogs.discovermagazine.com