Ming-Cong Su

Hello @yw0208, I am also curious about this work. May I ask how do you get 6D rotation from new_joints_vecs, i.e. what's the specific dimension in `(seq_len, 263)`? And have...

> > Hello @yw0208, I am also curious about this work. May I ask how do you get 6D rotation from new_joints_vecs, i.e. what's the specific dimension in `(seq_len, 263)`?...

Hi @EricGuo5513, thanks for replying to me. The shape of this data matrix (`HumanML3D\new_joint_vecs\007975.npy`) is `(84, 263)`. In my environment, the versions of numpy==1.21.6 and scipy==1.7.3 .

I reinstalled the `numpy==1.18.5` but the result is the same. I checked the result before running `motion_representation.ipynb` and producing `new_joint_vecs` and `new_joints`, which is the npy in `/joints`, but it...

> > 7975 > > > > Hi. Did you solve the issue? I also got `nan` for `007975.npy` data. Hi, I just skipped this data.

I met this problem when I built the environment on Windows. Then I download the dependency (SMPL model) in `prepare/download_smpl_model.sh` manually. (i.e. directly click on the link, download, and extract...

For everyone who meets this problem, I solve this by modifying the `SMPL_MODEL_DIR` in `config` as it's imported in `fit.py` `mgpt/data/transforms/joints2rots/config.py` line 115: 

Change the highlighted line into the correct path in your project directory. For example, changing line 115 into `SMPL_MODEL_DIR = "./deps/smpl_models"` works for me.

Hi @daidaiershidi, have you figured out why VQVAE+diffusion works badly on motion? Recently, I am also working on this scenario. (i.e. training VQVAE then use it to train latent diffusion...

Hi @Garfield-kh thanks for replying to me.😊 Can I ask how did you compute the FID for text2motion? it's because when I evaluated my own trained model, the value was...