A Deep Neural Network Tool for Automatic Segmentation of Human Body Parts in Natural Scenes
This work addresses the need for accurate body part segmentation in computer vision applications, but it appears incremental as it applies an existing method (Bayesian SegNet with concrete dropout) to a standard dataset.
The authors tackled the problem of automatically segmenting human body parts in natural scenes by training a Bayesian SegNet with concrete dropout on the Pascal-Parts dataset, achieving pixel-level predictions for categories like hair, head, and torso.
This short article describes a deep neural network trained to perform automatic segmentation of human body parts in natural scenes. More specifically, we trained a Bayesian SegNet with concrete dropout on the Pascal-Parts dataset to predict whether each pixel in a given frame was part of a person's hair, head, ear, eyebrows, legs, arms, mouth, neck, nose, or torso.