AWS Rekognition Human Emotions TrackBot

Evaluate Student and Patient Emotional Conditions using AWS Facial Recognition in Real Time

Comment 2

Inspiration

Under busy coursework, students are usually unaware of their learning performance and mental/emotional health. Also, it is difficult for instructors to identify students' understanding and struggle, especially in a very large lecture setting. Therefore, Amazon Machine Learning and Facial Recognition technology is here to help empower both instructors and students to realize their learning and emotional conditions in real time, and also over time!

What it does(Education)

  1. Instructors can take a picture of many students in the classroom anytime and AWS CompareFaces can accurately recognize a particular student face in the picture of many students.

  2. Then our program extracts the image of that particular student and initializes AWS Facial Detection to analyze the breakdowns of facial data, including age range, happiness/sadness, mouthOpen, and eyeOpen conditions.

  3. The powerful attributes are then processed into graphs to remind instructors and students about their leaning conditions which can be used to improve existing teaching and learning methods!

What it does(Medical)

In reality, it's usually hard for doctors to know the real mood state of patients, especially those with psychological/ neurological problems. However, knowing how the patient feel can be particularly beneficial in many ways. Researches have suggested that a good mood state can effectively help terminal disease sufferer to prolong their lives. Our program can:

  1. Correctly find the target patient living in a ward with several other patients and constantly record the mood state of target patient.
  2. Rely on Amazon Web Services to collect various data about the mood state of a particular patient and provide analysis to certain data depending on doctor's demand.
  3. Record the pattern of patients' daily lives. i.e. hour of sleep, time of eating. With our program, health care provider can:
  4. Quickly discover abnormal mood change among the huge data recorded.
  5. Understand how certain disease, treatment affect the mood state of a patient.
  6. Provide in-time mental help based on data.

How we built it

We use Amazon Web Service Facial ReKognition APIs using Java and at the end we use Google Chart to generate visualized graphs in real time.

Challenges we ran into

AWS Rekognition APIs packages could not be found or run successfully in certain IDEs. Also, we spent lots of time setting up the command line interface to run the APIs.

Accomplishments that we're proud of

The recognition and detection program can really accurately detect human faces and give great analysis of facial attributes! We're proud that our program now can empower both instructors and students!

What we learned

How to effectively use and integrate the powerful AWS development tools and APIs into our program!

What's next for AWS Rekognition Student Learning Tracker

Turn it into a more user-friendly and constructive app!

Share this project:

Updates