MCPcopy
hub / github.com/OpenMotionLab/MotionGPT / feats2joints

Method feats2joints

mGPT/data/Kit.py:48–52  ·  view source on GitHub ↗
(self, features)

Source from the content-addressed store, hash-verified

46 cfg.DATASET.NFEATS = self.nfeats
47
48 def feats2joints(self, features):
49 mean = torch.tensor(self.hparams.mean).to(features)
50 std = torch.tensor(self.hparams.std).to(features)
51 features = features * std + mean
52 return recover_from_ric(features, self.njoints)
53
54 def joints2feats(self, features):
55 features = process_file(features, self.njoints)[0]

Callers 8

load_motionFunction · 0.45
load_example_inputFunction · 0.45
forwardMethod · 0.45
val_t2m_forwardMethod · 0.45
val_m2m_forwardMethod · 0.45
train_vae_forwardMethod · 0.45
val_vae_forwardMethod · 0.45
save_npyMethod · 0.45

Calls 2

recover_from_ricFunction · 0.85
toMethod · 0.45

Tested by

no test coverage detected