When a human sees another human, they possess a wealth of information on how that body might exist within our physical world. When a machine sees a human in the same pose, the machine does not understand the pose and it’s implications. For example, the machine doesn’t know that my elbow bends inwards instead of outwards. For future robots interacting with humans, especially any robots that would be physically touching a human — it would be incredibly important for it to understand the ways in which I can and cannot bend.

This episode we speak with Christoph Lassner who worked on extracting a 3D model of a human pose & shape based on a single photo. We’ll show a few examples of what this looks like and how it works in the show.

Details: http://www.creativeai.net/posts/v8w2WYnQaahwcsbfs/smplify-3d-human-pose-and-shape-from-a-single-image

Follow the podcast: http://ai-guild.com

Want to see more Creative AI research? http://www.creativeai.net

Are you a Machine Learning Engineer looking for a new job? Machine Learning Jobs by Amazon: http://jobs.ai-guild.com