facebookresearch / facebookresearch/detectron2

How to distinguish between front of the head and back of the head from Densepose output

Open
#3,745 5 comments 2 reactions 0 assignees View on GitHub
densepose documentation
Dominant language
Python
Stars
34.7k
Forks
7.9k
PR merge metrics
No merged PRs in 30d

Description

## 📚 Documentation Issue

From the Densepose output, is there a way to know which head pixel belong the front of the head and which one belongs to back of the head? Especially, segmentation values 23 and 24 represent head segmentation, but I am not clear on which one is the front part of the head and which is the back part of the head. https://github.com/facebookresearch/DensePose/issues/64

On the similar track, is there a way to map UV values to identify eyes, nose of the human head in the original RGB input image?

Contributor guide

Open the contributing guide

Research direction

Start by reviewing the DensePose output semantics for segmentation values 23 and 24 and the referenced DensePose issue #64. Check whether the output provides a confirmed front/back head distinction or UV mapping for eyes and the nose; done means documenting the verified mapping or clearly stating its limitation.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision, documentation
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.