Pose Recommender
Envision
Updated on July 17, 2025
Important Note
If you are interested in this project, please contact me at wangshengwu01 [at] gmail [dot] com. We may collaborate on this project.
Many people struggle to capture photos that look natural, aesthetically pleasing, or express the intended mood. Amateur phtographers often cannot provide clear posing directions, and subjects frequently feel awkward or unsure how to stand or smile. Existing pose guide apps offer static catalogs of suggested poses, requiring manual browsing and imitation — they do not adapt to the actual model’s posture in real time.
We propose Pose Recommender, a framework for guiding users to pose themselves well that considers both the intrinsic user factors and the extrinsic environment factors in real time.
Product on the market
可颂 offers a feature called "灵感跟拍" that allows users to select an example photo and the app would detect the head of the model and guide the model to pose like the example photo. It is a good attempt to solve the problem, but it lacks the ability to understand the user's individually characteristic beauty and help them get the best pose that suits them and the environment.
Our idea
What we propose is an intelligent pose recommendation system. It understands the user's individual characteristics like face shape, body shape, hair style, and the environment like the background and the surrounding objects. Thus, it can recommend the best pose that suits the user and the environment.
Tools that may help
There are a couple of tools/models from TensorFlow.js that may help:
  • pose-detection (MoveNet, BlazePose, and PoseNet): detect the pose of the person(s) in real-time
  • body-segmentation: segment the person(s) and body parts in real-time
What we need to do
We need to build a model that can learn good poses from the professional models, influencers, and the users themselves. It is likely a semi-supervised learning model and it keeps learning from the users' feedback. It is a model that understands the relationship and interaction between the user and the environment.
Using cases/scenarios
When the user is standing by a book shelf, the model may suggest the user to take a book from the shelf and read it. To be more specific, it may propose how the user hold the book, like the angle of the book, the distance between the book and the user's face, the angle of the user's head, etc. It would also suggest the distance between the model and the photographer and the composition of the frame.
When the user is sitting by a table with a cup of coffee, the model may suggest the user to take a sip of the coffee. It may guide the model how to tilt the head, where the model should look at, and how to hold the cup.
It adapts accordingly to the user's intention, whether it is a daily photo or a professional portrait. That is to say, it would suggest different poses for different occasions.
© Wangsheng Wu