File size: 14,724 Bytes
1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 eff1e50 1399608 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 | """
Generic BVH → TopoSlots unified format converter.
Works for any BVH source: LAFAN1, Mixamo, Truebones Zoo, etc.
Produces the same Scheme C output as preprocess_humanml3d.py:
- skeleton.npz: skeleton graph
- motions/{id}.npz: per-motion features
- splits/all.txt: all motion IDs
- stats.npz: normalization stats
Usage:
python scripts/preprocess_bvh.py \
--bvh_dir data/raw/LAFAN1/bvh \
--output_dir data/processed/lafan1 \
--dataset_id lafan1 \
--target_fps 20 \
--remove_end_sites
"""
import sys
import argparse
from pathlib import Path
import numpy as np
from tqdm import tqdm
project_root = Path(__file__).parent.parent
sys.path.insert(0, str(project_root))
from src.data.bvh_parser import parse_bvh, resample_motion, remove_end_sites
from src.data.skeleton_graph import SkeletonGraph
from src.data.humanml3d_converter import _detect_foot_contact
def euler_to_6d_rotation(euler_angles: np.ndarray, order: str = 'ZYX') -> np.ndarray:
"""
Convert Euler angles (degrees) to continuous 6D rotation representation.
Uses scipy for correct BVH intrinsic Euler convention.
Args:
euler_angles: [..., 3] Euler angles in degrees
order: rotation order string (e.g., 'ZYX') — intrinsic convention
Returns:
[..., 6] continuous 6D rotation (first two columns of rotation matrix)
"""
from scipy.spatial.transform import Rotation
orig_shape = euler_angles.shape[:-1]
flat = euler_angles.reshape(-1, 3)
# BVH uses intrinsic rotations → scipy uppercase order
R = Rotation.from_euler(order.upper(), flat, degrees=True).as_matrix() # [N, 3, 3]
# Extract first two columns → 6D representation
rot_6d = np.concatenate([R[:, :, 0], R[:, :, 1]], axis=-1) # [N, 6]
return rot_6d.reshape(orig_shape + (6,)).astype(np.float32)
def forward_kinematics(
rotations: np.ndarray,
root_positions: np.ndarray,
offsets: np.ndarray,
parents: list[int],
rotation_order: str = 'ZYX',
local_translations: np.ndarray = None,
) -> np.ndarray:
"""
Compute global joint positions from local rotations + skeleton offsets via FK.
Uses scipy.spatial.transform.Rotation for correct BVH intrinsic Euler convention.
Verified against Blender's BVH FK (< 0.01mm error).
Args:
rotations: [T, J, 3] Euler angles in degrees (columns match rotation_order)
root_positions: [T, 3]
offsets: [J, 3] rest-pose offsets from parent (used when local_translations is None)
parents: [J] parent indices
rotation_order: Euler rotation order (e.g., 'ZYX') — intrinsic convention
local_translations: [T, J, 3] optional per-frame local translations
(for BVH files where all joints have position channels)
Returns:
[T, J, 3] global joint positions
"""
from scipy.spatial.transform import Rotation
T, J, _ = rotations.shape
positions = np.zeros((T, J, 3), dtype=np.float64)
global_rotmats = np.zeros((T, J, 3, 3), dtype=np.float64)
for j in range(J):
# Build local rotation matrix using scipy (intrinsic Euler)
local_rot = Rotation.from_euler(
rotation_order.upper(), rotations[:, j], degrees=True
).as_matrix() # [T, 3, 3]
p = parents[j]
if p < 0:
global_rotmats[:, j] = local_rot
positions[:, j] = root_positions
else:
global_rotmats[:, j] = np.einsum(
'tij,tjk->tik', global_rotmats[:, p], local_rot
)
# Use per-frame translations if available, otherwise static offsets
if local_translations is not None:
offset = local_translations[:, j, :] # [T, 3]
positions[:, j] = positions[:, p] + np.einsum(
'tij,tj->ti', global_rotmats[:, p], offset
)
else:
offset = offsets[j] # [3]
positions[:, j] = positions[:, p] + np.einsum(
'tij,j->ti', global_rotmats[:, p], offset
)
return positions.astype(np.float32)
def process_bvh_file(
bvh_path: Path,
target_fps: float,
max_frames: int,
min_frames: int,
do_remove_end_sites: bool,
manual_scale: float = None,
) -> dict | None:
"""Process a single BVH file into Scheme C format."""
try:
bvh = parse_bvh(bvh_path)
except Exception as e:
print(f" Failed to parse {bvh_path.name}: {e}")
return None
joint_names = bvh.skeleton.joint_names
parent_indices = bvh.skeleton.parent_indices
offsets = bvh.skeleton.rest_offsets
rotations = bvh.rotations
root_pos = bvh.root_positions
local_trans = bvh.local_translations # [T, J, 3] or None
# Remove end sites if requested
if do_remove_end_sites:
joint_names, parent_indices, offsets, rotations = remove_end_sites(
joint_names, parent_indices, offsets, rotations
)
# Also filter local_translations if present
if local_trans is not None:
keep_mask = [not name.endswith('_end') for name in bvh.skeleton.joint_names]
keep_indices = [i for i, k in enumerate(keep_mask) if k]
local_trans = local_trans[:, keep_indices, :]
# Remove dummy root: a static root joint whose only child is the real root (e.g. Hips).
if len(joint_names) > 1 and parent_indices[0] == -1:
children_of_root = [j for j in range(len(joint_names)) if parent_indices[j] == 0]
if len(children_of_root) == 1:
root_rot_range = rotations[:, 0].max(axis=0) - rotations[:, 0].min(axis=0)
root_is_static = np.all(root_rot_range < 1.0) # <1 degree range = static
if root_is_static:
old_root_name = joint_names[0]
child_idx = children_of_root[0]
# Use per-frame position of child as new root_pos if available
if local_trans is not None:
root_pos = local_trans[:, child_idx, :].copy()
local_trans = local_trans[:, 1:, :]
else:
root_pos = root_pos + offsets[child_idx]
# Remove joint 0
joint_names = joint_names[1:]
offsets = offsets[1:]
rotations = rotations[:, 1:]
# Remap parent indices
new_parents = []
for p in parent_indices[1:]:
if p <= 0:
new_parents.append(-1)
else:
new_parents.append(p - 1)
parent_indices = new_parents
print(f" Removed dummy root '{old_root_name}' → new root '{joint_names[0]}'")
J = len(joint_names)
# Resample to target FPS
if abs(bvh.fps - target_fps) > 0.5:
if local_trans is not None:
rotations, root_pos, local_trans = resample_motion(
rotations, root_pos, bvh.fps, target_fps, local_trans
)
else:
rotations, root_pos = resample_motion(
rotations, root_pos, bvh.fps, target_fps
)
T = rotations.shape[0]
if T < min_frames:
return None
if T > max_frames:
rotations = rotations[:max_frames]
root_pos = root_pos[:max_frames]
if local_trans is not None:
local_trans = local_trans[:max_frames]
T = max_frames
# Build skeleton graph
skeleton = SkeletonGraph(
joint_names=list(joint_names),
parent_indices=list(parent_indices),
rest_offsets=np.array(offsets, dtype=np.float32),
)
# Forward kinematics → global joint positions
joint_positions = forward_kinematics(
rotations, root_pos, offsets, parent_indices, bvh.rotation_order,
local_translations=local_trans,
)
# Scale normalization to meters
if manual_scale is not None:
scale = manual_scale
else:
# Auto-detect: BVH files commonly use centimeters
height = joint_positions[0, :, 1].max() - joint_positions[0, :, 1].min()
if height > 5.0: # almost certainly NOT meters (>5m body height)
scale = 0.01 # assume centimeters → meters
else:
scale = 1.0
if abs(scale - 1.0) > 0.001:
joint_positions = joint_positions * scale
root_pos = root_pos * scale
offsets = offsets * scale
# Center root at origin on first frame (XZ plane)
root_offset_xz = joint_positions[0, 0, [0, 2]].copy()
joint_positions[:, :, 0] -= root_offset_xz[0]
joint_positions[:, :, 2] -= root_offset_xz[1]
root_pos[:, 0] -= root_offset_xz[0]
root_pos[:, 2] -= root_offset_xz[1]
# Rebuild skeleton with scaled offsets
skeleton = SkeletonGraph(
joint_names=list(joint_names),
parent_indices=list(parent_indices),
rest_offsets=np.array(offsets, dtype=np.float32),
)
# Local positions (root-relative)
local_pos = joint_positions - joint_positions[:, 0:1, :]
# Velocities
vel = np.zeros_like(joint_positions)
vel[1:] = (joint_positions[1:] - joint_positions[:-1]) * target_fps
vel[0] = vel[1]
root_vel = vel[:, 0, :]
# Accelerations
acc = np.zeros_like(vel)
acc[1:] = (vel[1:] - vel[:-1]) * target_fps
acc[0] = acc[1]
# Bone lengths
bone_lengths = np.zeros((T, J), dtype=np.float32)
for j in range(J):
p = parent_indices[j]
if p >= 0:
bone_lengths[:, j] = np.linalg.norm(
joint_positions[:, j] - joint_positions[:, p], axis=-1
)
# Foot contact
foot_contact = _detect_foot_contact(joint_positions, vel, skeleton)
# 6D rotations (decoder GT)
rot_6d = euler_to_6d_rotation(rotations[:, 1:], bvh.rotation_order) # [T, J-1, 6]
return {
'skeleton': skeleton,
'data': {
'local_positions': local_pos.astype(np.float32),
'velocities': vel.astype(np.float32),
'root_position': root_pos.astype(np.float32),
'root_velocity': root_vel.astype(np.float32),
'joint_positions': joint_positions.astype(np.float32),
'local_rotations_6d': rot_6d.astype(np.float32),
'accelerations': acc.astype(np.float32),
'bone_lengths': bone_lengths.astype(np.float32),
'foot_contact': foot_contact.astype(np.float32),
'num_frames': T,
'fps': target_fps,
},
}
def preprocess_bvh_directory(
bvh_dir: str,
output_dir: str,
dataset_id: str,
target_fps: float = 20.0,
max_frames: int = 196,
min_frames: int = 24,
do_remove_end_sites: bool = True,
):
bvh_dir = Path(bvh_dir)
output_dir = Path(output_dir)
(output_dir / 'motions').mkdir(parents=True, exist_ok=True)
(output_dir / 'splits').mkdir(parents=True, exist_ok=True)
bvh_files = sorted(bvh_dir.rglob('*.bvh'))
print(f"Found {len(bvh_files)} BVH files in {bvh_dir}")
if not bvh_files:
print("No BVH files found, exiting.")
return
# Process all files
motion_ids = []
all_local_pos = []
all_velocities = []
first_skeleton = None
for i, bvh_path in enumerate(tqdm(bvh_files)):
result = process_bvh_file(
bvh_path, target_fps, max_frames, min_frames, do_remove_end_sites,
manual_scale=args.scale if hasattr(args, 'scale') else None,
)
if result is None:
continue
motion_id = f"{i:06d}"
skeleton = result['skeleton']
data = result['data']
if first_skeleton is None:
first_skeleton = skeleton
data['skeleton_id'] = dataset_id
data['texts'] = '' # No text for BVH data
data['source_file'] = bvh_path.name
np.savez_compressed(output_dir / 'motions' / f'{motion_id}.npz', **data)
motion_ids.append(motion_id)
# Collect stats (subsample)
if len(motion_ids) % 3 == 0:
all_local_pos.append(data['local_positions'])
all_velocities.append(data['velocities'])
print(f"\nProcessed: {len(motion_ids)} motions")
if not motion_ids:
print("No motions processed, exiting.")
return
# Save skeleton
np.savez(output_dir / 'skeleton.npz', **first_skeleton.to_dict())
print(f"Skeleton: {first_skeleton.num_joints} joints")
# Save stats
all_local_pos = np.concatenate(all_local_pos, axis=0)
all_velocities = np.concatenate(all_velocities, axis=0)
stats = {
'local_pos_mean': all_local_pos.mean(axis=0),
'local_pos_std': all_local_pos.std(axis=0) + 1e-8,
'velocity_mean': all_velocities.mean(axis=0),
'velocity_std': all_velocities.std(axis=0) + 1e-8,
'root_vel_mean': np.zeros(3, dtype=np.float32), # placeholder
'root_vel_std': np.ones(3, dtype=np.float32),
}
np.savez(output_dir / 'stats.npz', **stats)
# Save splits (80/10/10)
np.random.seed(42)
indices = np.random.permutation(len(motion_ids))
n_train = int(0.8 * len(indices))
n_val = int(0.1 * len(indices))
splits = {
'train': [motion_ids[i] for i in indices[:n_train]],
'val': [motion_ids[i] for i in indices[n_train:n_train + n_val]],
'test': [motion_ids[i] for i in indices[n_train + n_val:]],
'all': motion_ids,
}
for split_name, ids in splits.items():
with open(output_dir / 'splits' / f'{split_name}.txt', 'w') as f:
for mid in ids:
f.write(f'{mid}\n')
print(f" {split_name}: {len(ids)} motions")
print(f"\nDone! Output saved to {output_dir}")
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--bvh_dir', type=str, required=True)
parser.add_argument('--output_dir', type=str, required=True)
parser.add_argument('--dataset_id', type=str, required=True)
parser.add_argument('--target_fps', type=float, default=20.0)
parser.add_argument('--max_frames', type=int, default=196)
parser.add_argument('--min_frames', type=int, default=24)
parser.add_argument('--remove_end_sites', action='store_true')
parser.add_argument('--scale', type=float, default=None,
help='Manual scale factor (e.g., 0.01 for cm→m). Auto-detect if not set.')
args = parser.parse_args()
preprocess_bvh_directory(
args.bvh_dir, args.output_dir, args.dataset_id,
target_fps=args.target_fps,
max_frames=args.max_frames,
min_frames=args.min_frames,
do_remove_end_sites=args.remove_end_sites,
)
|