Jul 11, 2019 · In this paper, we observe that real and synthetic humans both have a skeleton (pose) representation. We found that the skeletons can effectively bridge the ...

Our proposed complementary learning technique learns a neural network model for multi-person part segmentation using a synthetic dataset and a real dataset.

Mar 5, 2021 · We presented a cross-domain complementary learning framework for multi-person part segmentation. Without using any real data part ...

We used publicly available Deep Neural Networks (DNNs) for feature labeling, specifically a Cross-Domain Complementary Learning (CDCL) DNN to segment different ...

This paper proposes a novel technique, called cross-domain complementary learning that takes advantage of the rich variations of real data and the easily ...

Our proposed comple- mentary learning technique learns a neural network model for multi-person part segmentation using a synthetic dataset and a real dataset.

1. CDCL+Pascal. 72.82. Cross-Domain Complementary Learning Using Pose for Multi-Person Part Segmentation ; 2. SCHP. 71.46. Self-Correction for Human Parsing.

Duration: 2:18
Posted: Aug 22, 2019

To make sure the synthetic data and real data are aligned in a common latent space, we use an auxiliary task of human pose estimation to bridge the two domains.

Multi-human parsing is an image segmentation task necessitating both instance-level and fine-grained category-level information.