MCPcopy Create free account
hub / github.com/ActiveVisionLab/DFNet / PoseLoss

Function PoseLoss

script/feature/misc.py:321–325  ·  view source on GitHub ↗
(args, pose_, pose, device)

Source from the content-addressed store, hash-verified

319 return torch.sqrt(torch.mean(torch.abs(input - target).pow(2)))
320
321def PoseLoss(args, pose_, pose, device):
322 loss_func = nn.MSELoss()
323 predict_pose = pose_.to(device) # maynot need reshape
324 pose_loss = loss_func(predict_pose, pose)
325 return pose_loss
326
327def to_img(x):
328 x = (x + 1.) * 0.5

Callers 2

train_on_batchFunction · 0.50

Calls

no outgoing calls

Tested by

no test coverage detected