I tested extd64 only got 5 fps
lucasjinreal opened this issue · comments
The model is tiny but speed is still not good though. It runs 0.4s per image, with some sample image from wider face.
Is that normal?
Thank you for the comment.
The images in wider face are mostly very large (largely over 1200x1200), and in that case, most face detectors including ours run slowly, so I think it can be possible.
Usually, the face detectors measure the speed using 320x320 or 640x640 images.
For more information, please refer the FPS graph in Figure 6 in our paper.
Thank you.
@yjyoo3312 Beside the size of image, does faces num effect inference speed? I mostly test on dense face image.
@jinfagang In evaluation, yes (because of NMS), but in demo setting, the face number does not largely affect the inference speed.
when i run the model it consumes ~3.5GB of memory. any idea why so?