search

bearpaw / PyraNet

Code for "Learning Feature Pyramids for Human Pose Estimation" (ICCV 2017)
Apache License 2.0
221 stars 58 forks source link

Realtime #1

Closed peixinx closed 7 years ago

peixinx commented 7 years ago

How is the realtime perfommance? It's said that this model needs 45.9GFLOPs per image. Does it works on videos?

bearpaw commented 7 years ago

Hi @peixinx . Our model is not real time. The test time of our final model (downloadable from Google Drive) is about 280 ms per image on a single Titan Black (6GB) GPU.