A tensorflow model of Posent by google is available here
The tensorflow team has also shared this blog post on posenet along with the github repo above. Now In the Blog post the author provide links to the two following papers:
- Towards Accurate Multi-person Pose Estimation in the Wild
- PersonLab: Person Pose Estimation and Instance Segmentation with a Bottom-Up, Part-Based, Geometric Embedding Model
In the post its not clearly mentioned which of the paper is the model based on so can you tell me which paper’s implementation is being used by the posenet model in Github repo above or the tflite version of posenet which is available here
By reading the github repo above I’m thinking that single pose detection is being done by the first paper and multi pose detection uses the second paper, is that correct? And which paper does the tflite version of posenet uses?
Can you please clear up these confusions.