search

mks0601 / I2L-MeshNet_RELEASE

Official PyTorch implementation of "I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image", ECCV 2020
MIT License
708 stars 130 forks source link

how do you test your inference time? #85

Closed KHao123 closed 2 years ago

KHao123 commented 2 years ago

I test your model with the speed_test in https://github.com/facebookresearch/LeViT/blob/main/speed_test.py. And get the result of image 300 fps on 2080ti and 0.42 fps on cpu, which is inconsistent with your 25 fps. Do you get the inference time on GPU or CPU？ Thank you.

mks0601 commented 2 years ago

The time is measured on a single RTX 2080Ti GPU with batch size 1.