Is this a supervised learning problem? If so, I was confused on how we obtain ground truth for body part regions? Is it hand-labeled data?
zhouwen
I think this is a supervised learning. To do this, they synthesized body part images. Original paper: https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/BodyPartRecognition.pdf
anon
In lecture, Kayvon mentioned using rendered scenes or motion capture data for ground truth (lecture video ~54:00).
xiaoruiL
Is there a standard for such classification of colors in terms of how to divide at the joints region?
Is this a supervised learning problem? If so, I was confused on how we obtain ground truth for body part regions? Is it hand-labeled data?
I think this is a supervised learning. To do this, they synthesized body part images. Original paper: https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/BodyPartRecognition.pdf
In lecture, Kayvon mentioned using rendered scenes or motion capture data for ground truth (lecture video ~54:00).
Is there a standard for such classification of colors in terms of how to divide at the joints region?