search

weigq / 3d_pose_baseline_pytorch

A simple baseline for 3d human pose estimation in PyTorch.
MIT License
370 stars 67 forks source link

human36m.zip does not exist #22

Open luzzou opened 5 years ago

luzzou commented 5 years ago

hi,

the link to download the pre-processed data human36m.zip does not exist, could you please upload it again?

AcKnight commented 5 years ago

hi,

I have the same problem with @Lu-Zou ,please upload it again, thx.

anibali commented 5 years ago

@weigq Could you please reupload the dataset or release code for transforming the .h5 files provided by https://github.com/una-dinosauria/3d-pose-baseline into the format used by this repo (including stat_3d.pth.tar)? At the moment it is not really possible to use your excellent work as-is since we can't train our own model or get the pretrained model.

wongkwan commented 5 years ago

I have the same problem, thank you so much If you can provide the data that pre-processed !

MCLYang commented 5 years ago

I just found the human3.6M but not sure where is the stat_3d.pth.tar https://github.com/facebookresearch/VideoPose3D/blob/master/DATASETS.md

agosztolai commented 4 years ago

Run this code to obtain the required input files stat_3d.pth.tar etc for the code. You'll need to have downloaded the h36m dataset first.


"""Utility functions for dealing with human3.6m data."""

from __future__ import division

import os
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import cameras
#import viz
import h5py
import glob
import copy
import torch

TRAIN_SUBJECTS = [1,5,6,7,8]
TEST_SUBJECTS  = [9,11]

# Joints in H3.6M -- data has 32 joints, but only 17 that move; these are the indices.
H36M_NAMES = ['']*32
H36M_NAMES[0]  = 'Hip'
H36M_NAMES[1]  = 'RHip'
H36M_NAMES[2]  = 'RKnee'
H36M_NAMES[3]  = 'RFoot'
H36M_NAMES[6]  = 'LHip'
H36M_NAMES[7]  = 'LKnee'
H36M_NAMES[8]  = 'LFoot'
H36M_NAMES[12] = 'Spine'
H36M_NAMES[13] = 'Thorax'
H36M_NAMES[14] = 'Neck/Nose'
H36M_NAMES[15] = 'Head'
H36M_NAMES[17] = 'LShoulder'
H36M_NAMES[18] = 'LElbow'
H36M_NAMES[19] = 'LWrist'
H36M_NAMES[25] = 'RShoulder'
H36M_NAMES[26] = 'RElbow'
H36M_NAMES[27] = 'RWrist'

# Stacked Hourglass produces 16 joints. These are the names.
SH_NAMES = ['']*16
SH_NAMES[0]  = 'RFoot'
SH_NAMES[1]  = 'RKnee'
SH_NAMES[2]  = 'RHip'
SH_NAMES[3]  = 'LHip'
SH_NAMES[4]  = 'LKnee'
SH_NAMES[5]  = 'LFoot'
SH_NAMES[6]  = 'Hip'
SH_NAMES[7]  = 'Spine'
SH_NAMES[8]  = 'Thorax'
SH_NAMES[9]  = 'Head'
SH_NAMES[10] = 'RWrist'
SH_NAMES[11] = 'RElbow'
SH_NAMES[12] = 'RShoulder'
SH_NAMES[13] = 'LShoulder'
SH_NAMES[14] = 'LElbow'
SH_NAMES[15] = 'LWrist'

def main():

    data_dir = '/data/h36m/'
    actions = define_actions('all')
    rcams = cameras.load_cameras(bpath=data_dir + '/cameras.h5')
    camera_frame = True #boolean. Whether to convert the data to camera coordinates

    # HG prediction (i.e. deeplabcut or similar) 
    train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use = \
    read_2d_predictions( actions, data_dir )

    torch.save(train_set, '/data/h36m/train_2d_ft.pth.tar')
    torch.save(test_set, '/data/h36m/test_2d_ft.pth.tar')

    #3D ground truth
    train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use = \
    create_2d_data( actions, data_dir, rcams )

    stat_2d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
#    torch.save(stat_2d, '/data/h36m/stat_2d.pth.tar')
    torch.save(train_set, '/data/h36m/train_2d.pth.tar')
    torch.save(test_set, '/data/h36m/test_2d.pth.tar')

    #3D ground truth
    train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, \
    train_root_positions, test_root_positions = \
    read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False )

    stat_3d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
    torch.save(stat_3d, '/data/h36m/stat_3d.pth.tar')
    torch.save(train_set, '/data/h36m/train_3d.pth.tar')
    torch.save(test_set, '/data/h36m/test_3d.pth.tar')

def load_data( bpath, subjects, actions, dim=3 ):
  """
  Loads 2d ground truth from disk, and puts it in an easy-to-acess dictionary

  Args
    bpath: String. Path where to load the data from
    subjects: List of integers. Subjects whose data will be loaded
    actions: List of strings. The actions to load
    dim: Integer={2,3}. Load 2 or 3-dimensional data
  Returns:
    data: Dictionary with keys k=(subject, action, seqname)
      values v=(nx(32*2) matrix of 2d ground truth)
      There will be 2 entries per subject/action if loading 3d data
      There will be 8 entries per subject/action if loading 2d data
  """

  if not dim in [2,3]:
    raise(ValueError, 'dim must be 2 or 3')

  data = {}

  for subj in subjects:
    for action in actions:

      print('Reading subject {0}, action {1}'.format(subj, action))

      dpath = os.path.join( bpath, 'S{0}'.format(subj), 'MyPoses/{0}D_positions'.format(dim), '{0}*.h5'.format(action) )
      print( dpath )

      fnames = glob.glob( dpath )

      loaded_seqs = 0
      for fname in fnames:
        seqname = os.path.basename( fname )

        # This rule makes sure SittingDown is not loaded when Sitting is requested
        if action == "Sitting" and seqname.startswith( "SittingDown" ):
          continue

        # This rule makes sure that WalkDog and WalkTogeter are not loaded when
        # Walking is requested.
        if seqname.startswith( action ):
          print( fname )
          loaded_seqs = loaded_seqs + 1

          with h5py.File( fname, 'r' ) as h5f:
            poses = h5f['{0}D_positions'.format(dim)][:]

          poses = poses.T
          data[ (subj, action, seqname) ] = poses

      if dim == 2:
        assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead".format( loaded_seqs )
      else:
        assert loaded_seqs == 2, "Expecting 2 sequences, found {0} instead".format( loaded_seqs )

  return data

def load_stacked_hourglass(data_dir, subjects, actions):
  """
  Load 2d detections from disk, and put it in an easy-to-acess dictionary.

  Args
    data_dir: string. Directory where to load the data from,
    subjects: list of integers. Subjects whose data will be loaded.
    actions: list of strings. The actions to load.
  Returns
    data: dictionary with keys k=(subject, action, seqname)
          values v=(nx(32*2) matrix of 2d stacked hourglass detections)
          There will be 2 entries per subject/action if loading 3d data
          There will be 8 entries per subject/action if loading 2d data
  """
  # Permutation that goes from SH detections to H36M ordering.
  SH_TO_GT_PERM = np.array([SH_NAMES.index( h ) for h in H36M_NAMES if h != '' and h in SH_NAMES])
  assert np.all( SH_TO_GT_PERM == np.array([6,2,1,0,3,4,5,7,8,9,13,14,15,12,11,10]) )

  data = {}

  for subj in subjects:
    for action in actions:

      print('Reading subject {0}, action {1}'.format(subj, action))

      dpath = os.path.join( data_dir, 'S{0}'.format(subj), 'StackedHourglass/{0}*.h5'.format(action) )
      print( dpath )

      fnames = glob.glob( dpath )

      loaded_seqs = 0
      for fname in fnames:
        seqname = os.path.basename( fname )
        seqname = seqname.replace('_',' ')

        # This rule makes sure SittingDown is not loaded when Sitting is requested
        if action == "Sitting" and seqname.startswith( "SittingDown" ):
          continue

        # This rule makes sure that WalkDog and WalkTogeter are not loaded when
        # Walking is requested.
        if seqname.startswith( action ):
          print( fname )
          loaded_seqs = loaded_seqs + 1

          # Load the poses from the .h5 file
          with h5py.File( fname, 'r' ) as h5f:
            poses = h5f['poses'][:]

            # Permute the loaded data to make it compatible with H36M
            poses = poses[:,SH_TO_GT_PERM,:]

            # Reshape into n x (32*2) matrix
            poses = np.reshape(poses,[poses.shape[0], -1])
            poses_final = np.zeros([poses.shape[0], len(H36M_NAMES)*2])

            dim_to_use_x    = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0] * 2
            dim_to_use_y    = dim_to_use_x+1

            dim_to_use = np.zeros(len(SH_NAMES)*2,dtype=np.int32)
            dim_to_use[0::2] = dim_to_use_x
            dim_to_use[1::2] = dim_to_use_y
            poses_final[:,dim_to_use] = poses
            seqname = seqname+'-sh'
            data[ (subj, action, seqname) ] = poses_final

      # Make sure we loaded 8 sequences
      if (subj == 11 and action == 'Directions'): # <-- this video is damaged
        assert loaded_seqs == 7, "Expecting 7 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )
      else:
        assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )

  return data

def normalization_stats(complete_data, dim, predict_14=False ):
  """
  Computes normalization statistics: mean and stdev, dimensions used and ignored

  Args
    complete_data: nxd np array with poses
    dim. integer={2,3} dimensionality of the data
    predict_14. boolean. Whether to use only 14 joints
  Returns
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dimensions_to_ignore: list of dimensions not used in the model
    dimensions_to_use: list of dimensions used in the model
  """
  if not dim in [2,3]:
    raise(ValueError, 'dim must be 2 or 3')

  data_mean = np.mean(complete_data, axis=0)
  data_std  =  np.std(complete_data, axis=0)

  # Encodes which 17 (or 14) 2d-3d pairs we are predicting
  dimensions_to_ignore = []
  if dim == 2:
    dimensions_to_use    = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0]
    dimensions_to_use    = np.sort( np.hstack( (dimensions_to_use*2, dimensions_to_use*2+1)))
    dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*2), dimensions_to_use )
  else: # dim == 3
    dimensions_to_use = np.where(np.array([x != '' for x in H36M_NAMES]))[0]
    dimensions_to_use = np.delete( dimensions_to_use, [0,7,9] if predict_14 else 0 )

    dimensions_to_use = np.sort( np.hstack( (dimensions_to_use*3,
                                             dimensions_to_use*3+1,
                                             dimensions_to_use*3+2)))
    dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*3), dimensions_to_use )

  return data_mean, data_std, dimensions_to_ignore, dimensions_to_use

def transform_world_to_camera(poses_set, cams, ncams=4 ):
    """
    Project 3d poses from world coordinate to camera coordinate system
    Args
      poses_set: dictionary with 3d poses
      cams: dictionary with cameras
      ncams: number of cameras per subject
    Return:
      t3d_camera: dictionary with 3d poses in camera coordinate
    """
    t3d_camera = {}
    for t3dk in sorted( poses_set.keys() ):

      subj, action, seqname = t3dk
      t3d_world = poses_set[ t3dk ]

      for c in range( ncams ):
        R, T, f, c, k, p, name = cams[ (subj, c+1) ]
        camera_coord = cameras.world_to_camera_frame( np.reshape(t3d_world, [-1, 3]), R, T)
        camera_coord = np.reshape( camera_coord, [-1, len(H36M_NAMES)*3] )

        sname = seqname[:-3]+"."+name+".h5" # e.g.: Waiting 1.58860488.h5
        t3d_camera[ (subj, action, sname) ] = camera_coord

    return t3d_camera

def normalize_data(data, data_mean, data_std, dim_to_use ):
  """
  Normalizes a dictionary of poses

  Args
    data: dictionary where values are
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dim_to_use: list of dimensions to keep in the data
  Returns
    data_out: dictionary with same keys as data, but values have been normalized
  """
  data_out = {}

  for key in data.keys():
    data[ key ] = data[ key ][ :, dim_to_use ]
    mu = data_mean[dim_to_use]
    stddev = data_std[dim_to_use]
    data_out[ key ] = np.divide( (data[key] - mu), stddev )

  return data_out

def unNormalizeData(normalized_data, data_mean, data_std, dimensions_to_ignore):
  """
  Un-normalizes a matrix whose mean has been substracted and that has been divided by
  standard deviation. Some dimensions might also be missing

  Args
    normalized_data: nxd matrix to unnormalize
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dimensions_to_ignore: list of dimensions that were removed from the original data
  Returns
    orig_data: the input normalized_data, but unnormalized
  """
  T = normalized_data.shape[0] # Batch size
  D = data_mean.shape[0] # Dimensionality

  orig_data = np.zeros((T, D), dtype=np.float32)
  dimensions_to_use = np.array([dim for dim in range(D)
                                if dim not in dimensions_to_ignore])

  orig_data[:, dimensions_to_use] = normalized_data

  # Multiply times stdev and add the mean
  stdMat = data_std.reshape((1, D))
  stdMat = np.repeat(stdMat, T, axis=0)
  meanMat = data_mean.reshape((1, D))
  meanMat = np.repeat(meanMat, T, axis=0)
  orig_data = np.multiply(orig_data, stdMat) + meanMat
  return orig_data

def define_actions( action ):
  """
  Given an action string, returns a list of corresponding actions.

  Args
    action: String. either "all" or one of the h36m actions
  Returns
    actions: List of strings. Actions to use.
  Raises
    ValueError: if the action is not a valid action in Human 3.6M
  """
  actions = ["Directions","Discussion","Eating","Greeting",
           "Phoning","Photo","Posing","Purchases",
           "Sitting","SittingDown","Smoking","Waiting",
           "WalkDog","Walking","WalkTogether"]

  if action == "All" or action == "all":
    return actions

  if not action in actions:
    raise( ValueError, "Unrecognized action: %s" % action )

  return [action]

def project_to_cameras( poses_set, cams, ncams=4 ):
  """
  Project 3d poses using camera parameters

  Args
    poses_set: dictionary with 3d poses
    cams: dictionary with camera parameters
    ncams: number of cameras per subject
  Returns
    t2d: dictionary with 2d poses
  """
  t2d = {}

  for t3dk in sorted( poses_set.keys() ):
    subj, a, seqname = t3dk
    t3d = poses_set[ t3dk ]

    for cam in range( ncams ):
      R, T, f, c, k, p, name = cams[ (subj, cam+1) ]
      pts2d, _, _, _, _ = cameras.project_point_radial( np.reshape(t3d, [-1, 3]), R, T, f, c, k, p )

      pts2d = np.reshape( pts2d, [-1, len(H36M_NAMES)*2] )
      sname = seqname[:-3]+"."+name+".h5" # e.g.: Waiting 1.58860488.h5
      t2d[ (subj, a, sname) ] = pts2d

  return t2d

def postprocess_3d( poses_set ):
  """
  Center 3d points around root

  Args
    poses_set: dictionary with 3d data
  Returns
    poses_set: dictionary with 3d data centred around root (center hip) joint
    root_positions: dictionary with the original 3d position of each pose
  """
  root_positions = {}
  for k in poses_set.keys():
    # Keep track of the global position
    root_positions[k] = copy.deepcopy(poses_set[k][:,:3])

    # Remove the root from the 3d position
    poses = poses_set[k]
    poses = poses - np.tile( poses[:,:3], [1, len(H36M_NAMES)] )
    poses_set[k] = poses

  return poses_set, root_positions

def read_2d_predictions( actions, data_dir ):
  """
  Loads 2d data from precomputed Stacked Hourglass detections

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
  Returns
    train_set: dictionary with loaded 2d stacked hourglass detections for training
    test_set: dictionary with loaded 2d stacked hourglass detections for testing
    data_mean: vector with the mean of the 2d training data
    data_std: vector with the standard deviation of the 2d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
  """

  train_set = load_stacked_hourglass( data_dir, TRAIN_SUBJECTS, actions)
  test_set  = load_stacked_hourglass( data_dir, TEST_SUBJECTS,  actions)

  complete_train = copy.deepcopy( np.vstack( train_set.values() ))
  data_mean, data_std,  dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

def create_2d_data( actions, data_dir, rcams ):
  """
  Creates 2d poses by projecting 3d poses with the corresponding camera
  parameters. Also normalizes the 2d poses

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
    rcams: dictionary with camera parameters
  Returns
    train_set: dictionary with projected 2d poses for training
    test_set: dictionary with projected 2d poses for testing
    data_mean: vector with the mean of the 2d training data
    data_std: vector with the standard deviation of the 2d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
  """

  # Load 3d data
  train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 )
  test_set  = load_data( data_dir, TEST_SUBJECTS,  actions, dim=3 )

  train_set = project_to_cameras( train_set, rcams )
  test_set  = project_to_cameras( test_set, rcams )

  # Compute normalization statistics.
  complete_train = copy.deepcopy( np.vstack( train_set.values() ))
  data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

  # Divide every dimension independently
  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

def read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False ):
  """
  Loads 3d poses, zero-centres and normalizes them

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
    camera_frame: boolean. Whether to convert the data to camera coordinates
    rcams: dictionary with camera parameters
    predict_14: boolean. Whether to predict only 14 joints
  Returns
    train_set: dictionary with loaded 3d poses for training
    test_set: dictionary with loaded 3d poses for testing
    data_mean: vector with the mean of the 3d training data
    data_std: vector with the standard deviation of the 3d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
    train_root_positions: dictionary with the 3d positions of the root in train
    test_root_positions: dictionary with the 3d positions of the root in test
  """
  # Load 3d data
  train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 )
  test_set  = load_data( data_dir, TEST_SUBJECTS,  actions, dim=3 )

  if camera_frame:
    train_set = transform_world_to_camera( train_set, rcams )
    test_set  = transform_world_to_camera( test_set, rcams )

  # Apply 3d post-processing (centering around root)
  train_set, train_root_positions = postprocess_3d( train_set )
  test_set,  test_root_positions  = postprocess_3d( test_set )

  # Compute normalization statistics
  complete_train = copy.deepcopy( np.vstack( train_set.values() ))
  data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=3, predict_14=predict_14 )

  # Divide every dimension independently
  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, train_root_positions, test_root_positions

if __name__ == "__main__":
    main()
raincrash commented 4 years ago

@agosztolai Where is the module named 'cameras' in the file you mentioned?

DuncanZauss commented 3 years ago

Hi @agosztolai and @raincrash , I created the .pth files with a slightly changed version of the code from @agosztolai .

  1. Clone this repo.
  2. Download the needed files from the human3.6m dataset as explained in https://github.com/una-dinosauria/3d-pose-baseline#first-of-all and save them in the data folder of this repo, also perform the subsequent renaming of the files and folders to have consistent names.
  3. Copy cameras.py from https://github.com/una-dinosauria/3d-pose-baseline/tree/master/src to the src folder of this repo.
  4. Finally run this code: 
    
"""Utility functions for dealing with human3.6m data."""

from future import division

import os import numpy as np import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D import src.cameras as cameras import cdflib

import viz

import h5py import glob import copy import torch

TRAIN_SUBJECTS = [1,5,6,7,8] TEST_SUBJECTS = [9,11]

Joints in H3.6M -- data has 32 joints, but only 17 that move; these are the indices.

H36M_NAMES = ['']*32 H36M_NAMES[0] = 'Hip' H36M_NAMES[1] = 'RHip' H36M_NAMES[2] = 'RKnee' H36M_NAMES[3] = 'RFoot' H36M_NAMES[6] = 'LHip' H36M_NAMES[7] = 'LKnee' H36M_NAMES[8] = 'LFoot' H36M_NAMES[12] = 'Spine' H36M_NAMES[13] = 'Thorax' H36M_NAMES[14] = 'Neck/Nose' H36M_NAMES[15] = 'Head' H36M_NAMES[17] = 'LShoulder' H36M_NAMES[18] = 'LElbow' H36M_NAMES[19] = 'LWrist' H36M_NAMES[25] = 'RShoulder' H36M_NAMES[26] = 'RElbow' H36M_NAMES[27] = 'RWrist'

Stacked Hourglass produces 16 joints. These are the names.

SH_NAMES = ['']*16 SH_NAMES[0] = 'RFoot' SH_NAMES[1] = 'RKnee' SH_NAMES[2] = 'RHip' SH_NAMES[3] = 'LHip' SH_NAMES[4] = 'LKnee' SH_NAMES[5] = 'LFoot' SH_NAMES[6] = 'Hip' SH_NAMES[7] = 'Spine' SH_NAMES[8] = 'Thorax' SH_NAMES[9] = 'Head' SH_NAMES[10] = 'RWrist' SH_NAMES[11] = 'RElbow' SH_NAMES[12] = 'RShoulder' SH_NAMES[13] = 'LShoulder' SH_NAMES[14] = 'LElbow' SH_NAMES[15] = 'LWrist'

def main(): data_dir = './data/h36m/' actions = define_actions('all') SUBJECT_IDS = [1,5,6,7,8,9,11] rcams = cameras.load_cameras("./data/h36m/metadata.xml", SUBJECT_IDS) camera_frame = True #boolean. Whether to convert the data to camera coordinates

#2D ground truth
train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use = \
create_2d_data( actions, data_dir, rcams )

stat_2d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
torch.save(stat_2d, './data/h36m/stat_2d.pth.tar')
torch.save(train_set, './data/h36m/train_2d.pth.tar')
torch.save(test_set, './data/h36m/test_2d.pth.tar')

#3D ground truth
train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, \
train_root_positions, test_root_positions = \
read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False )

stat_3d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
torch.save(stat_3d, './data/h36m/stat_3d.pth.tar')
torch.save(train_set, './data/h36m/train_3d.pth.tar')
torch.save(test_set, './data/h36m/test_3d.pth.tar')

def load_data( bpath, subjects, actions, dim=3 ): """Loads 2d ground truth from disk, and puts it in an easy-to-acess dictionary

Args bpath: String. Path where to load the data from subjects: List of integers. Subjects whose data will be loaded actions: List of strings. The actions to load dim: Integer={2,3}. Load 2 or 3-dimensional data Returns: data: Dictionary with keys k=(subject, action, seqname) values v=(nx(32*2) matrix of 2d ground truth) There will be 2 entries per subject/action if loading 3d data There will be 8 entries per subject/action if loading 2d data """

if not dim in [2,3]: raise ValueError('dim must be 2 or 3')

data = {}

for subj in subjects: for action in actions:

  print('Reading subject {0}, action {1}'.format(subj, action))

  dpath = os.path.join( bpath, 'S{0}'.format(subj), 'MyPoseFeatures/D{0}_Positions'.format(dim), '{0}*.cdf'.format(action) )
  print( dpath )

  fnames = glob.glob( dpath )

  loaded_seqs = 0
  for fname in fnames:
    seqname = os.path.basename( fname )

    # This rule makes sure SittingDown is not loaded when Sitting is requested
    if action == "Sitting" and seqname.startswith( "SittingDown" ):
      continue

    # This rule makes sure that WalkDog and WalkTogeter are not loaded when
    # Walking is requested.
    if seqname.startswith( action ):
      print( fname )
      loaded_seqs = loaded_seqs + 1

      cdf_file = cdflib.CDF(fname)
      poses = cdf_file.varget("Pose").squeeze()
      cdf_file.close()

      data[ (subj, action, seqname) ] = poses

  if dim == 2:
    assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead".format( loaded_seqs )
  else:
    assert loaded_seqs == 2, "Expecting 2 sequences, found {0} instead".format( loaded_seqs )

return data

def normalization_stats(complete_data, dim, predict_14=False ): """Computes normalization statistics: mean and stdev, dimensions used and ignored

Args complete_data: nxd np array with poses dim. integer={2,3} dimensionality of the data predict_14. boolean. Whether to use only 14 joints Returns data_mean: np vector with the mean of the data data_std: np vector with the standard deviation of the data dimensions_to_ignore: list of dimensions not used in the model dimensions_to_use: list of dimensions used in the model """ if not dim in [2,3]: raise ValueError('dim must be 2 or 3')

data_mean = np.mean(complete_data, axis=0) data_std = np.std(complete_data, axis=0)

Encodes which 17 (or 14) 2d-3d pairs we are predicting

dimensions_to_ignore = [] if dim == 2: dimensions_to_use = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0] dimensions_to_use = np.sort( np.hstack( (dimensions_to_use2, dimensions_to_use2+1))) dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*2), dimensions_to_use ) else: # dim == 3 dimensions_to_use = np.where(np.array([x != '' for x in H36M_NAMES]))[0] dimensions_to_use = np.delete( dimensions_to_use, [0,7,9] if predict_14 else 0 )

dimensions_to_use = np.sort( np.hstack( (dimensions_to_use*3,
                                         dimensions_to_use*3+1,
                                         dimensions_to_use*3+2)))
dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*3), dimensions_to_use )

return data_mean, data_std, dimensions_to_ignore, dimensions_to_use

def transform_world_to_camera(poses_set, cams, ncams=4 ): """Project 3d poses from world coordinate to camera coordinate system

Args
  poses_set: dictionary with 3d poses
  cams: dictionary with cameras
  ncams: number of cameras per subject
Return:
  t3d_camera: dictionary with 3d poses in camera coordinate
"""
t3d_camera = {}
for t3dk in sorted( poses_set.keys() ):

  subj, action, seqname = t3dk
  t3d_world = poses_set[ t3dk ]

  for c in range( ncams ):
    R, T, _, _, _, _, name = cams[ (subj, c+1) ]
    camera_coord = cameras.world_to_camera_frame( np.reshape(t3d_world, [-1, 3]), R, T)
    camera_coord = np.reshape( camera_coord, [-1, len(H36M_NAMES)*3] )

    sname = seqname[:-3]+ name + ".h5" # e.g.: Waiting 1.58860488.h5
    t3d_camera[ (subj, action, sname) ] = camera_coord

return t3d_camera

def normalize_data(data, data_mean, data_std, dim_to_use ): """Normalizes a dictionary of poses

Args data: dictionary where values are data_mean: np vector with the mean of the data data_std: np vector with the standard deviation of the data dim_to_use: list of dimensions to keep in the data Returns data_out: dictionary with same keys as data, but values have been normalized """ data_out = {}

for key in data.keys(): data[ key ] = data[ key ][ :, dim_to_use ] mu = data_mean[dim_to_use] stddev = data_std[dim_to_use] data_out[ key ] = np.divide( (data[key] - mu), stddev )

return data_out

def unNormalizeData(normalized_data, data_mean, data_std, dimensions_to_ignore): """Un-normalizes a matrix whose mean has been substracted and that has been divided by standard deviation. Some dimensions might also be missing

Args normalized_data: nxd matrix to unnormalize data_mean: np vector with the mean of the data data_std: np vector with the standard deviation of the data dimensions_to_ignore: list of dimensions that were removed from the original data Returns orig_data: the input normalized_data, but unnormalized """ T = normalized_data.shape[0] # Batch size D = data_mean.shape[0] # Dimensionality

orig_data = np.zeros((T, D), dtype=np.float32) dimensions_to_use = np.array([dim for dim in range(D) if dim not in dimensions_to_ignore])

orig_data[:, dimensions_to_use] = normalized_data

Multiply times stdev and add the mean

stdMat = data_std.reshape((1, D)) stdMat = np.repeat(stdMat, T, axis=0) meanMat = data_mean.reshape((1, D)) meanMat = np.repeat(meanMat, T, axis=0) orig_data = np.multiply(orig_data, stdMat) + meanMat return orig_data

def define_actions( action ): """Given an action string, returns a list of corresponding actions.

Args action: String. either "all" or one of the h36m actions Returns actions: List of strings. Actions to use. Raises ValueError: if the action is not a valid action in Human 3.6M """ actions = ["Directions","Discussion","Eating","Greeting", "Phoning","Photo","Posing","Purchases", "Sitting","SittingDown","Smoking","Waiting", "WalkDog","Walking","WalkTogether"]

if action == "All" or action == "all": return actions

if not action in actions: raise ValueError("Unrecognized action: %s" % action )

return [action]

def project_to_cameras( poses_set, cams, ncams=4 ): """ Project 3d poses using camera parameters

Args poses_set: dictionary with 3d poses cams: dictionary with camera parameters ncams: number of cameras per subject Returns t2d: dictionary with 2d poses """ t2d = {}

for t3dk in sorted( poses_set.keys() ): subj, a, seqname = t3dk t3d = poses_set[ t3dk ]

for cam in range( ncams ):
  R, T, f, c, k, p, name = cams[ (subj, cam+1) ]
  pts2d, _, _, _, _ = cameras.project_point_radial( np.reshape(t3d, [-1, 3]), R, T, f, c, k, p )

  pts2d = np.reshape( pts2d, [-1, len(H36M_NAMES)*2] )
  sname = seqname[:-3]+ name + ".h5"  # e.g.: Waiting 1.58860488.h5
  t2d[ (subj, a, sname) ] = pts2d

return t2d

def create_2d_data( actions, data_dir, rcams ): """Creates 2d poses by projecting 3d poses with the corresponding camera parameters. Also normalizes the 2d poses

Args actions: list of strings. Actions to load data_dir: string. Directory where the data can be loaded from rcams: dictionary with camera parameters Returns train_set: dictionary with projected 2d poses for training test_set: dictionary with projected 2d poses for testing data_mean: vector with the mean of the 2d training data data_std: vector with the standard deviation of the 2d training data dim_to_ignore: list with the dimensions to not predict dim_to_use: list with the dimensions to predict """

Load 3d data

train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 ) test_set = load_data( data_dir, TEST_SUBJECTS, actions, dim=3 )

Create 2d data by projecting with camera parameters

train_set = project_to_cameras( train_set, rcams ) test_set = project_to_cameras( test_set, rcams )

Compute normalization statistics.

complete_train = copy.deepcopy( np.vstack( list(train_set.values()) )) data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

Divide every dimension independently

train_set = normalize_data( train_set, data_mean, data_std, dim_to_use ) test_set = normalize_data( test_set, data_mean, data_std, dim_to_use )

return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

def read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False ): """Loads 3d poses, zero-centres and normalizes them

Args actions: list of strings. Actions to load data_dir: string. Directory where the data can be loaded from camera_frame: boolean. Whether to convert the data to camera coordinates rcams: dictionary with camera parameters predict_14: boolean. Whether to predict only 14 joints Returns train_set: dictionary with loaded 3d poses for training test_set: dictionary with loaded 3d poses for testing data_mean: vector with the mean of the 3d training data data_std: vector with the standard deviation of the 3d training data dim_to_ignore: list with the dimensions to not predict dim_to_use: list with the dimensions to predict train_root_positions: dictionary with the 3d positions of the root in train test_root_positions: dictionary with the 3d positions of the root in test """

Load 3d data

train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 ) test_set = load_data( data_dir, TEST_SUBJECTS, actions, dim=3 )

if camera_frame: train_set = transform_world_to_camera( train_set, rcams ) test_set = transform_world_to_camera( test_set, rcams )

Apply 3d post-processing (centering around root)

train_set, train_root_positions = postprocess_3d( train_set ) test_set, test_root_positions = postprocess_3d( test_set )

Compute normalization statistics

complete_train = copy.deepcopy( np.vstack( list(train_set.values()) )) data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=3, predict_14=predict_14 )

Divide every dimension independently

train_set = normalize_data( train_set, data_mean, data_std, dim_to_use ) test_set = normalize_data( test_set, data_mean, data_std, dim_to_use )

return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, train_root_positions, test_root_positions

def postprocess_3d( poses_set ): """Center 3d points around root

Args poses_set: dictionary with 3d data Returns poses_set: dictionary with 3d data centred around root (center hip) joint root_positions: dictionary with the original 3d position of each pose """ root_positions = {} for k in poses_set.keys():

Keep track of the global position

root_positions[k] = copy.deepcopy(poses_set[k][:,:3])

# Remove the root from the 3d position
poses = poses_set[k]
poses = poses - np.tile( poses[:,:3], [1, len(H36M_NAMES)] )
poses_set[k] = poses

return poses_set, root_positions

def load_stacked_hourglass(data_dir, subjects, actions): """ Load 2d detections from disk, and put it in an easy-to-acess dictionary.

Args data_dir: string. Directory where to load the data from, subjects: list of integers. Subjects whose data will be loaded. actions: list of strings. The actions to load. Returns data: dictionary with keys k=(subject, action, seqname) values v=(nx(32*2) matrix of 2d stacked hourglass detections) There will be 2 entries per subject/action if loading 3d data There will be 8 entries per subject/action if loading 2d data """

Permutation that goes from SH detections to H36M ordering.

SH_TO_GT_PERM = np.array([SH_NAMES.index( h ) for h in H36M_NAMES if h != '' and h in SH_NAMES]) assert np.all( SH_TO_GT_PERM == np.array([6,2,1,0,3,4,5,7,8,9,13,14,15,12,11,10]) )

data = {}

for subj in subjects: for action in actions:

  print('Reading subject {0}, action {1}'.format(subj, action))

  dpath = os.path.join( data_dir, 'S{0}'.format(subj), 'StackedHourglass/{0}*.h5'.format(action) )
  print( dpath )

  fnames = glob.glob( dpath )

  loaded_seqs = 0
  for fname in fnames:
    seqname = os.path.basename( fname )
    seqname = seqname.replace('_',' ')

    # This rule makes sure SittingDown is not loaded when Sitting is requested
    if action == "Sitting" and seqname.startswith( "SittingDown" ):
      continue

    # This rule makes sure that WalkDog and WalkTogeter are not loaded when
    # Walking is requested.
    if seqname.startswith( action ):
      print( fname )
      loaded_seqs = loaded_seqs + 1

      # Load the poses from the .h5 file
      with h5py.File( fname, 'r' ) as h5f:
        poses = h5f['poses'][:]

        # Permute the loaded data to make it compatible with H36M
        poses = poses[:,SH_TO_GT_PERM,:]

        # Reshape into n x (32*2) matrix
        poses = np.reshape(poses,[poses.shape[0], -1])
        poses_final = np.zeros([poses.shape[0], len(H36M_NAMES)*2])

        dim_to_use_x    = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0] * 2
        dim_to_use_y    = dim_to_use_x+1

        dim_to_use = np.zeros(len(SH_NAMES)*2,dtype=np.int32)
        dim_to_use[0::2] = dim_to_use_x
        dim_to_use[1::2] = dim_to_use_y
        poses_final[:,dim_to_use] = poses
        seqname = seqname+'-sh'
        data[ (subj, action, seqname) ] = poses_final

  # Make sure we loaded 8 sequences
  if (subj == 11 and action == 'Directions'): # <-- this video is damaged
    assert loaded_seqs == 7, "Expecting 7 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )
  else:
    assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )

return data

def read_2d_predictions( actions, data_dir ): """ Loads 2d data from precomputed Stacked Hourglass detections

Args actions: list of strings. Actions to load data_dir: string. Directory where the data can be loaded from Returns train_set: dictionary with loaded 2d stacked hourglass detections for training test_set: dictionary with loaded 2d stacked hourglass detections for testing data_mean: vector with the mean of the 2d training data data_std: vector with the standard deviation of the 2d training data dim_to_ignore: list with the dimensions to not predict dim_to_use: list with the dimensions to predict """

train_set = load_stacked_hourglass( data_dir, TRAIN_SUBJECTS, actions) test_set = load_stacked_hourglass( data_dir, TEST_SUBJECTS, actions)

complete_train = copy.deepcopy( np.vstack( train_set.values() )) data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

train_set = normalize_data( train_set, data_mean, data_std, dim_to_use ) test_set = normalize_data( test_set, data_mean, data_std, dim_to_use )

return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

if name == "main": main()


I hope this is helpful! :)
YuktiADY commented 1 year ago

Hi @agosztolai and @raincrash , I created the .pth files with a slightly changed version of the code from @agosztolai .

  1. Clone this repo.
  2. Download the needed files from the human3.6m dataset as explained in https://github.com/una-dinosauria/3d-pose-baseline#first-of-all and save them in the data folder of this repo, also perform the subsequent renaming of the files and folders to have consistent names.
  3. Copy cameras.py from https://github.com/una-dinosauria/3d-pose-baseline/tree/master/src to the src folder of this repo.
  4. Finally run this code:
"""Utility functions for dealing with human3.6m data."""

from __future__ import division

import os
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import src.cameras as cameras
import cdflib
#import viz
import h5py
import glob
import copy
import torch

TRAIN_SUBJECTS = [1,5,6,7,8]
TEST_SUBJECTS  = [9,11]

# Joints in H3.6M -- data has 32 joints, but only 17 that move; these are the indices.
H36M_NAMES = ['']*32
H36M_NAMES[0]  = 'Hip'
H36M_NAMES[1]  = 'RHip'
H36M_NAMES[2]  = 'RKnee'
H36M_NAMES[3]  = 'RFoot'
H36M_NAMES[6]  = 'LHip'
H36M_NAMES[7]  = 'LKnee'
H36M_NAMES[8]  = 'LFoot'
H36M_NAMES[12] = 'Spine'
H36M_NAMES[13] = 'Thorax'
H36M_NAMES[14] = 'Neck/Nose'
H36M_NAMES[15] = 'Head'
H36M_NAMES[17] = 'LShoulder'
H36M_NAMES[18] = 'LElbow'
H36M_NAMES[19] = 'LWrist'
H36M_NAMES[25] = 'RShoulder'
H36M_NAMES[26] = 'RElbow'
H36M_NAMES[27] = 'RWrist'

# Stacked Hourglass produces 16 joints. These are the names.
SH_NAMES = ['']*16
SH_NAMES[0]  = 'RFoot'
SH_NAMES[1]  = 'RKnee'
SH_NAMES[2]  = 'RHip'
SH_NAMES[3]  = 'LHip'
SH_NAMES[4]  = 'LKnee'
SH_NAMES[5]  = 'LFoot'
SH_NAMES[6]  = 'Hip'
SH_NAMES[7]  = 'Spine'
SH_NAMES[8]  = 'Thorax'
SH_NAMES[9]  = 'Head'
SH_NAMES[10] = 'RWrist'
SH_NAMES[11] = 'RElbow'
SH_NAMES[12] = 'RShoulder'
SH_NAMES[13] = 'LShoulder'
SH_NAMES[14] = 'LElbow'
SH_NAMES[15] = 'LWrist'

def main():
    data_dir = './data/h36m/'
    actions = define_actions('all')
    SUBJECT_IDS = [1,5,6,7,8,9,11]
    rcams = cameras.load_cameras("./data/h36m/metadata.xml", SUBJECT_IDS)
    camera_frame = True #boolean. Whether to convert the data to camera coordinates

    #2D ground truth
    train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use = \
    create_2d_data( actions, data_dir, rcams )

    stat_2d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
    torch.save(stat_2d, './data/h36m/stat_2d.pth.tar')
    torch.save(train_set, './data/h36m/train_2d.pth.tar')
    torch.save(test_set, './data/h36m/test_2d.pth.tar')

    #3D ground truth
    train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, \
    train_root_positions, test_root_positions = \
    read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False )

    stat_3d = {'mean': data_mean, 'std': data_std, 'dim_use': dim_to_use, 'dim_ignore': dim_to_ignore}
    torch.save(stat_3d, './data/h36m/stat_3d.pth.tar')
    torch.save(train_set, './data/h36m/train_3d.pth.tar')
    torch.save(test_set, './data/h36m/test_3d.pth.tar')

def load_data( bpath, subjects, actions, dim=3 ):
  """Loads 2d ground truth from disk, and puts it in an easy-to-acess dictionary

  Args
    bpath: String. Path where to load the data from
    subjects: List of integers. Subjects whose data will be loaded
    actions: List of strings. The actions to load
    dim: Integer={2,3}. Load 2 or 3-dimensional data
  Returns:
    data: Dictionary with keys k=(subject, action, seqname)
      values v=(nx(32*2) matrix of 2d ground truth)
      There will be 2 entries per subject/action if loading 3d data
      There will be 8 entries per subject/action if loading 2d data
  """

  if not dim in [2,3]:
    raise ValueError('dim must be 2 or 3')

  data = {}

  for subj in subjects:
    for action in actions:

      print('Reading subject {0}, action {1}'.format(subj, action))

      dpath = os.path.join( bpath, 'S{0}'.format(subj), 'MyPoseFeatures/D{0}_Positions'.format(dim), '{0}*.cdf'.format(action) )
      print( dpath )

      fnames = glob.glob( dpath )

      loaded_seqs = 0
      for fname in fnames:
        seqname = os.path.basename( fname )

        # This rule makes sure SittingDown is not loaded when Sitting is requested
        if action == "Sitting" and seqname.startswith( "SittingDown" ):
          continue

        # This rule makes sure that WalkDog and WalkTogeter are not loaded when
        # Walking is requested.
        if seqname.startswith( action ):
          print( fname )
          loaded_seqs = loaded_seqs + 1

          cdf_file = cdflib.CDF(fname)
          poses = cdf_file.varget("Pose").squeeze()
          cdf_file.close()

          data[ (subj, action, seqname) ] = poses

      if dim == 2:
        assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead".format( loaded_seqs )
      else:
        assert loaded_seqs == 2, "Expecting 2 sequences, found {0} instead".format( loaded_seqs )

  return data

def normalization_stats(complete_data, dim, predict_14=False ):
  """Computes normalization statistics: mean and stdev, dimensions used and ignored

  Args
    complete_data: nxd np array with poses
    dim. integer={2,3} dimensionality of the data
    predict_14. boolean. Whether to use only 14 joints
  Returns
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dimensions_to_ignore: list of dimensions not used in the model
    dimensions_to_use: list of dimensions used in the model
  """
  if not dim in [2,3]:
    raise ValueError('dim must be 2 or 3')

  data_mean = np.mean(complete_data, axis=0)
  data_std  =  np.std(complete_data, axis=0)

  # Encodes which 17 (or 14) 2d-3d pairs we are predicting
  dimensions_to_ignore = []
  if dim == 2:
    dimensions_to_use    = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0]
    dimensions_to_use    = np.sort( np.hstack( (dimensions_to_use*2, dimensions_to_use*2+1)))
    dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*2), dimensions_to_use )
  else: # dim == 3
    dimensions_to_use = np.where(np.array([x != '' for x in H36M_NAMES]))[0]
    dimensions_to_use = np.delete( dimensions_to_use, [0,7,9] if predict_14 else 0 )

    dimensions_to_use = np.sort( np.hstack( (dimensions_to_use*3,
                                             dimensions_to_use*3+1,
                                             dimensions_to_use*3+2)))
    dimensions_to_ignore = np.delete( np.arange(len(H36M_NAMES)*3), dimensions_to_use )

  return data_mean, data_std, dimensions_to_ignore, dimensions_to_use

def transform_world_to_camera(poses_set, cams, ncams=4 ):
    """Project 3d poses from world coordinate to camera coordinate system

    Args
      poses_set: dictionary with 3d poses
      cams: dictionary with cameras
      ncams: number of cameras per subject
    Return:
      t3d_camera: dictionary with 3d poses in camera coordinate
    """
    t3d_camera = {}
    for t3dk in sorted( poses_set.keys() ):

      subj, action, seqname = t3dk
      t3d_world = poses_set[ t3dk ]

      for c in range( ncams ):
        R, T, _, _, _, _, name = cams[ (subj, c+1) ]
        camera_coord = cameras.world_to_camera_frame( np.reshape(t3d_world, [-1, 3]), R, T)
        camera_coord = np.reshape( camera_coord, [-1, len(H36M_NAMES)*3] )

        sname = seqname[:-3]+ name + ".h5" # e.g.: Waiting 1.58860488.h5
        t3d_camera[ (subj, action, sname) ] = camera_coord

    return t3d_camera

def normalize_data(data, data_mean, data_std, dim_to_use ):
  """Normalizes a dictionary of poses

  Args
    data: dictionary where values are
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dim_to_use: list of dimensions to keep in the data
  Returns
    data_out: dictionary with same keys as data, but values have been normalized
  """
  data_out = {}

  for key in data.keys():
    data[ key ] = data[ key ][ :, dim_to_use ]
    mu = data_mean[dim_to_use]
    stddev = data_std[dim_to_use]
    data_out[ key ] = np.divide( (data[key] - mu), stddev )

  return data_out

def unNormalizeData(normalized_data, data_mean, data_std, dimensions_to_ignore):
  """Un-normalizes a matrix whose mean has been substracted and that has been divided by
  standard deviation. Some dimensions might also be missing

  Args
    normalized_data: nxd matrix to unnormalize
    data_mean: np vector with the mean of the data
    data_std: np vector with the standard deviation of the data
    dimensions_to_ignore: list of dimensions that were removed from the original data
  Returns
    orig_data: the input normalized_data, but unnormalized
  """
  T = normalized_data.shape[0] # Batch size
  D = data_mean.shape[0] # Dimensionality

  orig_data = np.zeros((T, D), dtype=np.float32)
  dimensions_to_use = np.array([dim for dim in range(D)
                                if dim not in dimensions_to_ignore])

  orig_data[:, dimensions_to_use] = normalized_data

  # Multiply times stdev and add the mean
  stdMat = data_std.reshape((1, D))
  stdMat = np.repeat(stdMat, T, axis=0)
  meanMat = data_mean.reshape((1, D))
  meanMat = np.repeat(meanMat, T, axis=0)
  orig_data = np.multiply(orig_data, stdMat) + meanMat
  return orig_data

def define_actions( action ):
  """Given an action string, returns a list of corresponding actions.

  Args
    action: String. either "all" or one of the h36m actions
  Returns
    actions: List of strings. Actions to use.
  Raises
    ValueError: if the action is not a valid action in Human 3.6M
  """
  actions = ["Directions","Discussion","Eating","Greeting",
           "Phoning","Photo","Posing","Purchases",
           "Sitting","SittingDown","Smoking","Waiting",
           "WalkDog","Walking","WalkTogether"]

  if action == "All" or action == "all":
    return actions

  if not action in actions:
    raise ValueError("Unrecognized action: %s" % action )

  return [action]

def project_to_cameras( poses_set, cams, ncams=4 ):
  """
  Project 3d poses using camera parameters

  Args
    poses_set: dictionary with 3d poses
    cams: dictionary with camera parameters
    ncams: number of cameras per subject
  Returns
    t2d: dictionary with 2d poses
  """
  t2d = {}

  for t3dk in sorted( poses_set.keys() ):
    subj, a, seqname = t3dk
    t3d = poses_set[ t3dk ]

    for cam in range( ncams ):
      R, T, f, c, k, p, name = cams[ (subj, cam+1) ]
      pts2d, _, _, _, _ = cameras.project_point_radial( np.reshape(t3d, [-1, 3]), R, T, f, c, k, p )

      pts2d = np.reshape( pts2d, [-1, len(H36M_NAMES)*2] )
      sname = seqname[:-3]+ name + ".h5"  # e.g.: Waiting 1.58860488.h5
      t2d[ (subj, a, sname) ] = pts2d

  return t2d

def create_2d_data( actions, data_dir, rcams ):
  """Creates 2d poses by projecting 3d poses with the corresponding camera
  parameters. Also normalizes the 2d poses

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
    rcams: dictionary with camera parameters
  Returns
    train_set: dictionary with projected 2d poses for training
    test_set: dictionary with projected 2d poses for testing
    data_mean: vector with the mean of the 2d training data
    data_std: vector with the standard deviation of the 2d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
  """

  # Load 3d data
  train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 )
  test_set  = load_data( data_dir, TEST_SUBJECTS,  actions, dim=3 )

  # Create 2d data by projecting with camera parameters
  train_set = project_to_cameras( train_set, rcams )
  test_set  = project_to_cameras( test_set, rcams )

  # Compute normalization statistics.
  complete_train = copy.deepcopy( np.vstack( list(train_set.values()) ))
  data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

  # Divide every dimension independently
  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

def read_3d_data( actions, data_dir, camera_frame, rcams, predict_14=False ):
  """Loads 3d poses, zero-centres and normalizes them

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
    camera_frame: boolean. Whether to convert the data to camera coordinates
    rcams: dictionary with camera parameters
    predict_14: boolean. Whether to predict only 14 joints
  Returns
    train_set: dictionary with loaded 3d poses for training
    test_set: dictionary with loaded 3d poses for testing
    data_mean: vector with the mean of the 3d training data
    data_std: vector with the standard deviation of the 3d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
    train_root_positions: dictionary with the 3d positions of the root in train
    test_root_positions: dictionary with the 3d positions of the root in test
  """
  # Load 3d data
  train_set = load_data( data_dir, TRAIN_SUBJECTS, actions, dim=3 )
  test_set  = load_data( data_dir, TEST_SUBJECTS,  actions, dim=3 )

  if camera_frame:
    train_set = transform_world_to_camera( train_set, rcams )
    test_set  = transform_world_to_camera( test_set, rcams )

  # Apply 3d post-processing (centering around root)
  train_set, train_root_positions = postprocess_3d( train_set )
  test_set,  test_root_positions  = postprocess_3d( test_set )

  # Compute normalization statistics
  complete_train = copy.deepcopy( np.vstack( list(train_set.values()) ))
  data_mean, data_std, dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=3, predict_14=predict_14 )

  # Divide every dimension independently
  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use, train_root_positions, test_root_positions

def postprocess_3d( poses_set ):
  """Center 3d points around root

  Args
    poses_set: dictionary with 3d data
  Returns
    poses_set: dictionary with 3d data centred around root (center hip) joint
    root_positions: dictionary with the original 3d position of each pose
  """
  root_positions = {}
  for k in poses_set.keys():
    # Keep track of the global position
    root_positions[k] = copy.deepcopy(poses_set[k][:,:3])

    # Remove the root from the 3d position
    poses = poses_set[k]
    poses = poses - np.tile( poses[:,:3], [1, len(H36M_NAMES)] )
    poses_set[k] = poses

  return poses_set, root_positions

def load_stacked_hourglass(data_dir, subjects, actions):
  """
  Load 2d detections from disk, and put it in an easy-to-acess dictionary.

  Args
    data_dir: string. Directory where to load the data from,
    subjects: list of integers. Subjects whose data will be loaded.
    actions: list of strings. The actions to load.
  Returns
    data: dictionary with keys k=(subject, action, seqname)
          values v=(nx(32*2) matrix of 2d stacked hourglass detections)
          There will be 2 entries per subject/action if loading 3d data
          There will be 8 entries per subject/action if loading 2d data
  """
  # Permutation that goes from SH detections to H36M ordering.
  SH_TO_GT_PERM = np.array([SH_NAMES.index( h ) for h in H36M_NAMES if h != '' and h in SH_NAMES])
  assert np.all( SH_TO_GT_PERM == np.array([6,2,1,0,3,4,5,7,8,9,13,14,15,12,11,10]) )

  data = {}

  for subj in subjects:
    for action in actions:

      print('Reading subject {0}, action {1}'.format(subj, action))

      dpath = os.path.join( data_dir, 'S{0}'.format(subj), 'StackedHourglass/{0}*.h5'.format(action) )
      print( dpath )

      fnames = glob.glob( dpath )

      loaded_seqs = 0
      for fname in fnames:
        seqname = os.path.basename( fname )
        seqname = seqname.replace('_',' ')

        # This rule makes sure SittingDown is not loaded when Sitting is requested
        if action == "Sitting" and seqname.startswith( "SittingDown" ):
          continue

        # This rule makes sure that WalkDog and WalkTogeter are not loaded when
        # Walking is requested.
        if seqname.startswith( action ):
          print( fname )
          loaded_seqs = loaded_seqs + 1

          # Load the poses from the .h5 file
          with h5py.File( fname, 'r' ) as h5f:
            poses = h5f['poses'][:]

            # Permute the loaded data to make it compatible with H36M
            poses = poses[:,SH_TO_GT_PERM,:]

            # Reshape into n x (32*2) matrix
            poses = np.reshape(poses,[poses.shape[0], -1])
            poses_final = np.zeros([poses.shape[0], len(H36M_NAMES)*2])

            dim_to_use_x    = np.where(np.array([x != '' and x != 'Neck/Nose' for x in H36M_NAMES]))[0] * 2
            dim_to_use_y    = dim_to_use_x+1

            dim_to_use = np.zeros(len(SH_NAMES)*2,dtype=np.int32)
            dim_to_use[0::2] = dim_to_use_x
            dim_to_use[1::2] = dim_to_use_y
            poses_final[:,dim_to_use] = poses
            seqname = seqname+'-sh'
            data[ (subj, action, seqname) ] = poses_final

      # Make sure we loaded 8 sequences
      if (subj == 11 and action == 'Directions'): # <-- this video is damaged
        assert loaded_seqs == 7, "Expecting 7 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )
      else:
        assert loaded_seqs == 8, "Expecting 8 sequences, found {0} instead. S:{1} {2}".format(loaded_seqs, subj, action )

  return data

def read_2d_predictions( actions, data_dir ):
  """
  Loads 2d data from precomputed Stacked Hourglass detections

  Args
    actions: list of strings. Actions to load
    data_dir: string. Directory where the data can be loaded from
  Returns
    train_set: dictionary with loaded 2d stacked hourglass detections for training
    test_set: dictionary with loaded 2d stacked hourglass detections for testing
    data_mean: vector with the mean of the 2d training data
    data_std: vector with the standard deviation of the 2d training data
    dim_to_ignore: list with the dimensions to not predict
    dim_to_use: list with the dimensions to predict
  """

  train_set = load_stacked_hourglass( data_dir, TRAIN_SUBJECTS, actions)
  test_set  = load_stacked_hourglass( data_dir, TEST_SUBJECTS,  actions)

  complete_train = copy.deepcopy( np.vstack( train_set.values() ))
  data_mean, data_std,  dim_to_ignore, dim_to_use = normalization_stats( complete_train, dim=2 )

  train_set = normalize_data( train_set, data_mean, data_std, dim_to_use )
  test_set  = normalize_data( test_set,  data_mean, data_std, dim_to_use )

  return train_set, test_set, data_mean, data_std, dim_to_ignore, dim_to_use

if __name__ == "__main__":
    main()

I hope this is helpful! :)

After downloading all the D positions . What renaming we need to do.

which file to run after this. ???

Can you explain ?