OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields

1. Background: Top-Down vs. Bottom-Up Approaches


2. OpenPose’s Core Innovation: Part Affinity Fields (PAFs)

PAFs solve the grouping problem by encoding both location and orientation of limbs:


3. Architecture

  1. Backbone: Truncated VGG-19 (first 10 layers) for feature extraction.
  2. Two Parallel Branches:
  3. Multi-Stage Refinement:

4. Greedy Parsing with PAFs

1. Detect Joint Candidates

2. Score Possible Limb Connections

For each limb type (e.g., neck-to-hip):