Spaces:
Running
on
Zero
Running
on
Zero
File size: 11,283 Bytes
1ea89dd |
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 |
import json
import os
from collections import defaultdict
from typing import Any, Dict, List, Optional
import numpy as np
import torch
import torch.distributed as dist
import torch.utils.data.distributed
import wandb
from PIL import Image
from torch.nn import functional as F
from torch.utils.data import DataLoader
from tqdm import tqdm
from unik3d.utils.distributed import barrier, get_world_size, is_main_process
from unik3d.utils.misc import remove_leading_dim, remove_padding, ssi_helper
from unik3d.utils.visualization import colorize, image_grid
def stack_mixedshape_numpy(tensor_list, dim=0):
max_rows = max(tensor.shape[0] for tensor in tensor_list)
max_columns = max(tensor.shape[1] for tensor in tensor_list)
padded_tensors = []
for tensor in tensor_list:
rows, columns, *_ = tensor.shape
pad_rows = max_rows - rows
pad_columns = max_columns - columns
padded_tensor = np.pad(
tensor, ((0, pad_rows), (0, pad_columns), (0, 0)), mode="constant"
)
padded_tensors.append(padded_tensor)
return np.stack(padded_tensors, axis=dim)
def original_image(batch):
paddings = [
torch.tensor(pads)
for img_meta in batch["img_metas"]
for pads in img_meta.get("paddings", [[0] * 4])
]
paddings = torch.stack(paddings).to(batch["data"]["image"].device)[
..., [0, 2, 1, 3]
] # lrtb
T, _, H, W = batch["data"]["depth"].shape
batch["data"]["image"] = F.interpolate(
batch["data"]["image"],
(H + paddings[0][2] + paddings[0][3], W + paddings[0][1] + paddings[0][2]),
mode="bilinear",
align_corners=False,
antialias=True,
)
batch["data"]["image"] = remove_padding(
batch["data"]["image"], paddings.repeat(T, 1)
)
return batch
def original_image_inv(batch, preds=None):
paddings = [
torch.tensor(pads)
for img_meta in batch["img_metas"]
for pads in img_meta.get("padding_size", [[0] * 4])
]
T, _, H, W = batch["data"]["depth"].shape
batch["data"]["image"] = remove_padding(batch["data"]["image"], paddings * T)
batch["data"]["image"] = F.interpolate(
batch["data"]["image"],
(H, W),
mode="bilinear",
align_corners=False,
antialias=True,
)
if preds is not None:
for key in ["depth"]:
if key in preds:
preds[key] = remove_padding(preds[key], paddings * T)
preds[key] = F.interpolate(
preds[key],
(H, W),
mode="bilinear",
align_corners=False,
antialias=True,
)
return batch, preds
def aggregate_metrics(metrics_all, exclude_fn=lambda name: False):
aggregate_name = "".join(
[name_ds[:3] for name_ds in metrics_all.keys() if not exclude_fn(name_ds)]
)
metrics_aggregate = defaultdict(list)
for name_ds, metrics in metrics_all.items():
if exclude_fn(name_ds):
continue
for metrics_name, metrics_value in metrics.items():
metrics_aggregate[metrics_name].append(metrics_value)
return {
**{aggregate_name: {k: sum(v) / len(v) for k, v in metrics_aggregate.items()}},
**metrics_all,
}
GROUPS = {
"SFoV": ["KITTI", "NYUv2Depth", "DiodeIndoor", "ETH3D", "IBims"],
"SFoVDi": ["DiodeIndoor_F", "ETH3D_F", "IBims_F"],
"LFoV": ["ADT", "KITTI360", "ScanNetpp_F"],
}
def aggregate_metrics_camera(metrics_all):
available_groups = {
k: v for k, v in GROUPS.items() if any([name in metrics_all for name in v])
}
for group_name, group_datasets in available_groups.items():
metrics_aggregate = defaultdict(list)
for dataset_name in group_datasets:
if dataset_name not in metrics_all:
print(
f"Dataset {dataset_name} not used for aggregation of {group_name}"
)
continue
for metrics_name, metrics_value in metrics_all[dataset_name].items():
metrics_aggregate[metrics_name].append(metrics_value)
metrics_all[group_name] = {
k: sum(v) / len(v) for k, v in metrics_aggregate.items()
}
return metrics_all
def log_metrics(metrics_all, step):
for name_ds, metrics in metrics_all.items():
for metrics_name, metrics_value in metrics.items():
try:
wandb.log(
{f"Metrics/{name_ds}/{metrics_name}": metrics_value}, step=step
)
except:
print(f"Metrics/{name_ds}/{metrics_name} {round(metrics_value, 4)}")
def log_artifacts(artifacts_all, step, run_id):
for ds_name, artifacts in artifacts_all.items():
rgbs, gts = artifacts["rgbs"], artifacts["gts"]
logging_imgs = [
*rgbs,
*gts,
*[
x
for k, v in artifacts.items()
if ("rgbs" not in k and "gts" not in k)
for x in v
],
]
artifacts_grid = image_grid(logging_imgs, len(artifacts), len(rgbs))
try:
wandb.log({f"{ds_name}_test": [wandb.Image(artifacts_grid)]}, step=step)
except:
print(f"Error while saving artifacts at step {step}")
def show(vals, dataset, ssi_depth=False):
output_artifacts, additionals = {}, {}
predictions, gts, errors, images = [], [], [], []
for v in vals:
image = v["image"][0].unsqueeze(0)
gt = v["depth"][0].unsqueeze(0)
prediction = v["depth_pred"][0].unsqueeze(0)
# Downsample for memory and viz
# if any([x in dataset.__class__.__name__ for x in ["DDAD", "Argoverse", "Waymo", "DrivingStereo"]]):
# gt = F.interpolate(gt, scale_factor=0.5, mode="nearest-exact")
# # Dilate for a better visualization
# gt[gt < 1e-4] = dilate(gt)[gt < 1e-4]
H, W = gt.shape[-2:]
aspect_ratio = H / W
new_W = int((300_000 / aspect_ratio) ** 0.5)
new_H = int(aspect_ratio * new_W)
gt = F.interpolate(gt, (new_H, new_W), mode="nearest-exact")
# Format predictions and errors for every metrics used
prediction = F.interpolate(
prediction,
gt.shape[-2:],
mode="bilinear",
align_corners=False,
antialias=True,
)
error = torch.zeros_like(prediction)
error[gt > dataset.min_depth] = (
4
* dataset.max_depth
* torch.abs(gt - prediction)[gt > dataset.min_depth]
/ gt[gt > dataset.min_depth]
)
if ssi_depth:
scale, shift = ssi_helper(gt[gt > 0], prediction[gt > 0])
prediction = (prediction * scale + shift).clip(0.0, dataset.max_depth)
prediction = colorize(
prediction.squeeze().cpu().detach().numpy(),
vmin=dataset.min_depth,
vmax=dataset.max_depth,
cmap="magma_r",
)
error = error.clip(0.0, dataset.max_depth).cpu().detach().numpy()
error = colorize(error.squeeze(), vmin=0.001, vmax=1.0, cmap="coolwarm")
errors.append(error)
predictions.append(prediction)
image = F.interpolate(
image, gt.shape[-2:], mode="bilinear", align_corners=False, antialias=True
)
image = image.cpu().detach() * dataset.normalization_stats["std"].view(
1, -1, 1, 1
) + dataset.normalization_stats["mean"].view(1, -1, 1, 1)
image = (
(255 * image)
.clip(0.0, 255.0)
.to(torch.uint8)
.permute(0, 2, 3, 1)
.numpy()
.squeeze()
)
gt = gt.clip(0.0, dataset.max_depth).cpu().detach().numpy()
gt = colorize(
gt.squeeze(), vmin=dataset.min_depth, vmax=dataset.max_depth, cmap="magma_r"
)
gts.append(gt)
images.append(image)
for name, additional in v.get("infos", {}).items():
if name not in additionals:
additionals[name] = []
if additional[0].shape[0] == 3:
val = (
(127.5 * (additional[0] + 1))
.clip(0, 255)
.to(torch.uint8)
.cpu()
.detach()
.permute(1, 2, 0)
.numpy()
)
else:
val = colorize(
additional[0].cpu().detach().squeeze().numpy(),
0.0,
dataset.max_depth,
)
additionals[name].append(val)
output_artifacts.update(
{
f"predictions": stack_mixedshape_numpy(predictions),
f"errors": stack_mixedshape_numpy(errors),
"rgbs": stack_mixedshape_numpy(images),
"gts": stack_mixedshape_numpy(gts),
**{k: stack_mixedshape_numpy(v) for k, v in additionals.items()},
}
)
return output_artifacts
METRIC_B = "F1"
INVERT = True
SSI_VISUALIZATION = True
def validate(
model,
test_loaders: Dict[str, DataLoader],
step,
run_id,
context,
idxs=(1, 100, 150, 1000),
):
metrics_all, predictions_select = {}, {}
world_size = get_world_size()
for name_ds, test_loader in test_loaders.items():
idxs = [idx % len(test_loader.dataset) for idx in idxs]
ds_show = []
for i, batch in enumerate(test_loader):
with context:
batch["data"] = {
k: v.to(model.device) for k, v in batch["data"].items()
}
preds = model(batch["data"], batch["img_metas"])
if batch["data"]["image"].ndim == 5:
batch["data"] = remove_leading_dim(batch["data"])
if preds["depth"].ndim == 5:
preds = remove_leading_dim(preds)
batch = original_image(batch)
test_loader.dataset.accumulate_metrics(
inputs=batch["data"],
preds=preds,
keyframe_idx=batch["img_metas"][0].get("keyframe_idx"),
)
# for prediction images logging
if i * world_size in idxs:
ii = (len(preds["depth"]) + 1) // 2 - 1
slice_ = slice(ii, ii + 1)
batch["data"] = {k: v[slice_] for k, v in batch["data"].items()}
preds["depth"] = preds["depth"][slice_]
ds_show.append({**batch["data"], **{"depth_pred": preds["depth"]}})
barrier()
metrics_all[name_ds] = test_loader.dataset.get_evaluation()
predictions_select[name_ds] = show(
ds_show, test_loader.dataset, ssi_depth=SSI_VISUALIZATION
)
barrier()
if is_main_process():
log_artifacts(artifacts_all=predictions_select, step=step, run_id=run_id)
metrics_all = aggregate_metrics(
metrics_all, exclude_fn=lambda name: "mono" in name
)
metrics_all = aggregate_metrics_camera(metrics_all)
log_metrics(metrics_all=metrics_all, step=step)
return metrics_all
|