Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Hotfix/sg 000 fix sigmas #1503

Merged
merged 5 commits into from
Oct 5, 2023
Merged

Hotfix/sg 000 fix sigmas #1503

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
97b15e6
Fixed a bug where OKS sigmas had two incorrect values
BloodAxe Oct 5, 2023
6b2b59c
Fixed a bug where OKS sigmas had two incorrect values
BloodAxe Oct 5, 2023
622a196
Allow user setting the SUPER_GRADIENTS_COCO_DATASET_DIR env variable …
BloodAxe Oct 5, 2023
38b5f29
Remove excess stuff
BloodAxe Oct 5, 2023
e553b71
Merge branch 'master' into hotfix/SG-000-fix-sigmas
BloodAxe Oct 5, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion documentation/source/PoseEstimation.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -284,7 +284,7 @@ Then, the full configuration file should look like this:
num_joints: 17

# OKs sigma values take from https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocotools/cocoeval.py#L523
oks_sigmas: [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 1.007, 1.007, 0.087, 0.087, 0.089, 0.089]
oks_sigmas: [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 0.107, 0.107, 0.087, 0.087, 0.089, 0.089]

train_dataset_params:
image_paths: /my_new_dataset/train/images
Expand Down
72 changes: 72 additions & 0 deletions src/super_gradients/recipes/dataset_params/coco_pose_estimation_common_dataset_params.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,72 @@
# This is not "true" dataset params, one cannot use it to instantiate dataloaders
# But it contains skeleton definitions for COCO2017 dataset and exists to avoid
# duplication of those parameters in other dataset params

num_joints: 17

# OKs sigma values take from https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocotools/cocoeval.py#L523
oks_sigmas: [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 0.107, 0.107, 0.087, 0.087, 0.089, 0.089]

flip_indexes: [ 0, 2, 1, 4, 3, 6, 5, 8, 7, 10, 9, 12, 11, 14, 13, 16, 15,]

edge_links:
- [0, 1]
- [0, 2]
- [1, 2]
- [1, 3]
- [2, 4]
- [3, 5]
- [4, 6]
- [5, 6]
- [5, 7]
- [5, 11]
- [6, 8]
- [6, 12]
- [7, 9]
- [8, 10]
- [11, 12]
- [11, 13]
- [12, 14]
- [13, 15]
- [14, 16]

edge_colors:
- [214, 39, 40] # Nose -> LeftEye
- [148, 103, 189] # Nose -> RightEye
- [44, 160, 44] # LeftEye -> RightEye
- [140, 86, 75] # LeftEye -> LeftEar
- [227, 119, 194] # RightEye -> RightEar
- [127, 127, 127] # LeftEar -> LeftShoulder
- [188, 189, 34] # RightEar -> RightShoulder
- [127, 127, 127] # Shoulders
- [188, 189, 34] # LeftShoulder -> LeftElbow
- [140, 86, 75] # LeftTorso
- [23, 190, 207] # RightShoulder -> RightElbow
- [227, 119, 194] # RightTorso
- [31, 119, 180] # LeftElbow -> LeftArm
- [255, 127, 14] # RightElbow -> RightArm
- [148, 103, 189] # Waist
- [255, 127, 14] # Left Hip -> Left Knee
- [214, 39, 40] # Right Hip -> Right Knee
- [31, 119, 180] # Left Knee -> Left Ankle
- [44, 160, 44] # Right Knee -> Right Ankle


keypoint_colors:
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
71 changes: 3 additions & 68 deletions src/super_gradients/recipes/dataset_params/coco_pose_estimation_dataset_params.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,71 +1,6 @@
num_joints: 17

# OKs sigma values take from https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocotools/cocoeval.py#L523
oks_sigmas: [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 1.007, 1.007, 0.087, 0.087, 0.089, 0.089]

flip_indexes: [ 0, 2, 1, 4, 3, 6, 5, 8, 7, 10, 9, 12, 11, 14, 13, 16, 15,]

edge_links:
- [0, 1]
- [0, 2]
- [1, 2]
- [1, 3]
- [2, 4]
- [3, 5]
- [4, 6]
- [5, 6]
- [5, 7]
- [5, 11]
- [6, 8]
- [6, 12]
- [7, 9]
- [8, 10]
- [11, 12]
- [11, 13]
- [12, 14]
- [13, 15]
- [14, 16]

edge_colors:
- [214, 39, 40] # Nose -> LeftEye
- [148, 103, 189] # Nose -> RightEye
- [44, 160, 44] # LeftEye -> RightEye
- [140, 86, 75] # LeftEye -> LeftEar
- [227, 119, 194] # RightEye -> RightEar
- [127, 127, 127] # LeftEar -> LeftShoulder
- [188, 189, 34] # RightEar -> RightShoulder
- [127, 127, 127] # Shoulders
- [188, 189, 34] # LeftShoulder -> LeftElbow
- [140, 86, 75] # LeftTorso
- [23, 190, 207] # RightShoulder -> RightElbow
- [227, 119, 194] # RightTorso
- [31, 119, 180] # LeftElbow -> LeftArm
- [255, 127, 14] # RightElbow -> RightArm
- [148, 103, 189] # Waist
- [255, 127, 14] # Left Hip -> Left Knee
- [214, 39, 40] # Right Hip -> Right Knee
- [31, 119, 180] # Left Knee -> Left Ankle
- [44, 160, 44] # Right Knee -> Right Ankle


keypoint_colors:
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
- [31, 119, 180]
- [148, 103, 189]
defaults:
- coco_pose_estimation_common_dataset_params
- _self_


train_dataset_params:
Expand Down
29 changes: 3 additions & 26 deletions ...super_gradients/recipes/dataset_params/coco_pose_estimation_rescoring_dataset_params.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,29 +1,6 @@
num_joints: 17

# OKs sigma values take from https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocotools/cocoeval.py#L523
oks_sigmas: [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 1.007, 1.007, 0.087, 0.087, 0.089, 0.089]

edge_links:
- [0, 1]
- [0, 2]
- [1, 2]
- [1, 3]
- [2, 4]
- [3, 5]
- [4, 6]
- [5, 6]
- [5, 7]
- [5, 11]
- [6, 8]
- [6, 12]
- [7, 9]
- [8, 10]
- [11, 12]
- [11, 13]
- [12, 14]
- [13, 15]
- [14, 16]

defaults:
- coco_pose_estimation_common_dataset_params
- _self_

train_dataset_params:
pkl_file: REPLACE_ME_WITH_PATH_TO_TRAIN_DATASET.PKL
Expand Down
2 changes: 1 addition & 1 deletion src/super_gradients/training/metrics/pose_estimation_metrics.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,7 +105,7 @@ def __init__(

if oks_sigmas is None:
if num_joints == 17:
oks_sigmas = np.array([0.26, 0.25, 0.25, 0.35, 0.35, 0.79, 0.79, 0.72, 0.72, 0.62, 0.62, 1.07, 1.07, 0.87, 0.87, 0.89, 0.89]) / 10.0
oks_sigmas = np.array([0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 0.107, 0.107, 0.087, 0.087, 0.089, 0.089])
else:
oks_sigmas = np.array([0.1] * num_joints)
logger.warning(
Expand Down
8 changes: 7 additions & 1 deletion tests/integration_tests/pose_estimation_models_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
import os
import unittest

import torch
Expand All @@ -14,14 +15,17 @@

class PoseEstimationModelsIntegrationTest(unittest.TestCase):
def setUp(self):
self.oks_sigmas = [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 1.007, 1.007, 0.087, 0.087, 0.089, 0.089]
self.oks_sigmas = [0.026, 0.025, 0.025, 0.035, 0.035, 0.079, 0.079, 0.072, 0.072, 0.062, 0.062, 0.107, 0.107, 0.087, 0.087, 0.089, 0.089]
self.flip_indexes = [0, 2, 1, 4, 3, 6, 5, 8, 7, 10, 9, 12, 11, 14, 13, 16, 15]
# This is for easy testing on local machine - you can set this environment variable to your own COCO dataset location
self.data_dir = os.environ.get("SUPER_GRADIENTS_COCO_DATASET_DIR", "/data/coco")

def test_dekr_model(self):
val_loader = get_data_loader(
"coco_pose_estimation_dekr_dataset_params",
COCOKeypointsDataset,
train=False,
dataset_params=dict(data_dir=self.data_dir),
dataloader_params=dict(num_workers=0),
)

Expand Down Expand Up @@ -54,6 +58,7 @@ def test_dekr_model_with_tta(self):
"coco_pose_estimation_dekr_dataset_params",
COCOKeypointsDataset,
train=False,
dataset_params=dict(data_dir=self.data_dir),
dataloader_params=dict(num_workers=0),
)

Expand Down Expand Up @@ -84,6 +89,7 @@ def test_dekr_model_with_rescoring(self):
"coco_pose_estimation_dekr_dataset_params",
COCOKeypointsDataset,
train=False,
dataset_params=dict(data_dir=self.data_dir),
dataloader_params=dict(batch_size=1, num_workers=0),
)

Expand Down