sjsufall2019-cmpe272 / virtual-yoga-trainer Goto Github PK

Since data is key to effectively train any ML model, how are you planning to gather the required data for this project. Are you using any readily available dataset online, or would you gather your own data and label them accordingly.

Will the app be able to spot exactly where the posture need to be improved, and provide any feedback on how it could be improved. Will the app be able to provide a feedback by highlighting a specific region in the user photo so that they could have a reference on what exactly needs to be done to improve the posture.

When the user uses a high resolution camera to capture the image, the image would have high density pixels, which would result in an increase in the time taken to train your ML model. Will you be using any technique to reduce this overhead.

It happens sometimes that a single picture might not be adequate to analyze the yoga posture and confirm if it’s the right form or not. It might require more data to further validate and provide a feedback with higher accuracy to the user. So, how exactly will the user data be provided to the ML mode, will you be providing a stream of images as an input or a single image to detect the accuracy of the posture.

Hey guys,
Just trying to test out your app, I can get it to run and am able to login, but I think the poses are missing from my clean db... any chance you can share those? as currently going to the poses route gives a page with just the header... any help or direction would be great...

By categorizing the yoga poses based on the difficulty levels like: low medium and high would increase customer engagement and usability factor of the application resulting in increased business value