Search results

NME calculation for 68 points

Status: Open.

#29 In protossw512/AdaptiveWingLoss;

· hyperparameters opened on Sep 13, 2021

what does the model return if there is no face in the image?

Status: Open.

#28 In protossw512/AdaptiveWingLoss;

· segalinc opened on Sep 8, 2021

Get 4.68 NME on WFLW dataset

Status: Open.

#27 In protossw512/AdaptiveWingLoss;

· Suke0 opened on Aug 25, 2021

Two years have passed, have you considered release the train code?

Status: Open.

#26 In protossw512/AdaptiveWingLoss;

· BeierZhu opened on Jun 7, 2021

When will the pretrained model on 300W be released?

Status: Open.

#25 In protossw512/AdaptiveWingLoss;

· jimmysue opened on May 25, 2020

Please consider releasing the training code

Status: Open.

#24 In protossw512/AdaptiveWingLoss;

· noranart opened on Apr 30, 2020

Run the code on custom data?

Status: Open.

#23 In protossw512/AdaptiveWingLoss;

· antithing opened on Apr 29, 2020

why facial landmarks for covered parts?

Status: Open.

#22 In protossw512/AdaptiveWingLoss;

· sri9s opened on Apr 24, 2020

recently can u realese the training code

Status: Open.

#21 In protossw512/AdaptiveWingLoss;

· jiangtaoo2333 opened on Apr 24, 2020

68 Facial landmarks model

Status: Open.

#16 In protossw512/AdaptiveWingLoss;

· dguari1 opened on Mar 5, 2020

Training code implementation

Status: Open.

#14 In protossw512/AdaptiveWingLoss;

· HassanAbbas92 opened on Feb 18, 2020

Why does the output have one more channel than the num of landmarks?

Status: Open.

#13 In protossw512/AdaptiveWingLoss;

· ty2806 opened on Feb 13, 2020