Revert commits df08552 and b94f6a6.

Pose2Sim/Demo_Batch/Config.toml

@@ -32,24 +32,19 @@ frame_range = [] # For example [10,300], or [] for all frames.
 exclude_from_batch = [] # List of trials to be excluded from batch analysis, ['<participant_dir/trial_dir>', 'etc'].
 # e.g. ['S00_P00_Participant/S00_P00_T00_StaticTrial', 'S00_P00_Participant/S00_P00_T01_BalancingTrial']
 
-
 [pose]
 vid_img_extension = 'mp4' # any video or image extension
-
-pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 model), Whole_body (COCO_133: body + feet + hands), Body (COCO_17)
+pose_model = 'HALPE_26'  #With RTMLib: HALPE_26 (body and feet, default), COCO_133 (body, feet, hands), COCO_17 (body)
                          # /!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
                          #With MMPose: HALPE_26, COCO_133, COCO_17, CUSTOM. See CUSTOM example at the end of the file
                          #With openpose: BODY_25B, BODY_25, BODY_135, COCO, MPII
                          #With mediapipe: BLAZEPOSE
                          #With alphapose: HALPE_26, HALPE_68, HALPE_136, COCO_133
                          #With deeplabcut: CUSTOM. See example at the end of the file
 mode = 'balanced' # 'lightweight', 'balanced', 'performance'
-det_frequency = 1 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
+det_frequency = 100 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
                   # Equal to or greater than 1, can be as high as you want in simple uncrowded cases. Much faster, but might be less accurate. 
-device = 'auto' # 'auto', 'CPU', 'CUDA', 'MPS', 'ROCM'
-backend = 'auto' # 'auto', 'openvino', 'onnxruntime', 'opencv'
-
-display_detection = true
+display_detection = false
 overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 save_video = 'to_video' # 'to_video' or 'to_images', 'none', or ['to_video', 'to_images']
 output_format = 'openpose' # 'openpose', 'mmpose', 'deeplabcut', 'none' or a list of them # /!\ only 'openpose' is supported for now

Pose2Sim/Demo_Batch/Trial_1/Config.toml

@@ -32,12 +32,10 @@
 # exclude_from_batch = [] # List of trials to be excluded from batch analysis, ['<participant_dir/trial_dir>', 'etc'].
 # # e.g. ['S00_P00_Participant/S00_P00_T00_StaticTrial', 'S00_P00_Participant/S00_P00_T01_BalancingTrial']
 
-
 # [pose]
 # vid_img_extension = 'mp4' # any video or image extension
-
-# pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 model), Whole_body (COCO_133: body + feet + hands), Body (COCO_17)
-                         # # /!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
+# pose_model = 'HALPE_26'  #With RTMLib: HALPE_26 (body and feet, default), COCO_133 (body, feet, hands), COCO_17 (body)
+                         # #/!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
                          # #With MMPose: HALPE_26, COCO_133, COCO_17, CUSTOM. See CUSTOM example at the end of the file
                          # #With openpose: BODY_25B, BODY_25, BODY_135, COCO, MPII
                          # #With mediapipe: BLAZEPOSE
@@ -46,11 +44,8 @@
 # mode = 'balanced' # 'lightweight', 'balanced', 'performance'
 # det_frequency = 100 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
                   # # Equal to or greater than 1, can be as high as you want in simple uncrowded cases. Much faster, but might be less accurate. 
-# device = 'auto' # 'auto', 'CPU', 'CUDA', 'MPS', 'ROCM'
-# backend = 'auto' # 'auto', 'openvino', 'onnxruntime', 'opencv'
-
+# # overwrite_pose = true # set to false if you don't want to recalculate pose estimation when it has already been done
 # display_detection = true
-# overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 # save_video = 'to_video' # 'to_video' or 'to_images', 'none', or ['to_video', 'to_images']
 # output_format = 'openpose' # 'openpose', 'mmpose', 'deeplabcut', 'none' or a list of them # /!\ only 'openpose' is supported for now
 

Pose2Sim/Demo_Batch/Trial_2/Config.toml

@@ -34,9 +34,8 @@ participant_mass = [70.0, 63.5] # kg # Only used for marker augmentation and sca
 
 # [pose]
 # vid_img_extension = 'mp4' # any video or image extension
-
-# pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 model), Whole_body (COCO_133: body + feet + hands), Body (COCO_17)
-                         # # /!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
+# pose_model = 'HALPE_26'  #With RTMLib: HALPE_26 (body and feet, default), COCO_133 (body, feet, hands), COCO_17 (body)
+                         # #/!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
                          # #With MMPose: HALPE_26, COCO_133, COCO_17, CUSTOM. See CUSTOM example at the end of the file
                          # #With openpose: BODY_25B, BODY_25, BODY_135, COCO, MPII
                          # #With mediapipe: BLAZEPOSE
@@ -45,11 +44,8 @@ participant_mass = [70.0, 63.5] # kg # Only used for marker augmentation and sca
 # mode = 'balanced' # 'lightweight', 'balanced', 'performance'
 # det_frequency = 100 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
                   # # Equal to or greater than 1, can be as high as you want in simple uncrowded cases. Much faster, but might be less accurate. 
-# device = 'auto' # 'auto', 'CPU', 'CUDA', 'MPS', 'ROCM'
-# backend = 'auto' # 'auto', 'openvino', 'onnxruntime', 'opencv'
-
+# # overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 # display_detection = true
-# overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 # save_video = 'to_video' # 'to_video' or 'to_images', 'none', or ['to_video', 'to_images']
 # output_format = 'openpose' # 'openpose', 'mmpose', 'deeplabcut', 'none' or a list of them # /!\ only 'openpose' is supported for now
 

Pose2Sim/Demo_MultiPerson/Config.toml

@@ -32,11 +32,9 @@ frame_range = [] # For example [10,300], or [] for all frames.
 exclude_from_batch = [] # List of trials to be excluded from batch analysis, ['<participant_dir/trial_dir>', 'etc'].
 # e.g. ['S00_P00_Participant/S00_P00_T00_StaticTrial', 'S00_P00_Participant/S00_P00_T01_BalancingTrial']
 
-
 [pose]
 vid_img_extension = 'mp4' # any video or image extension
-
-pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 model), Whole_body (COCO_133: body + feet + hands), Body (COCO_17)
+pose_model = 'HALPE_26'  #With RTMLib: HALPE_26 (body and feet, default), COCO_133 (body, feet, hands), COCO_17 (body)
                          # /!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
                          #With MMPose: HALPE_26, COCO_133, COCO_17, CUSTOM. See CUSTOM example at the end of the file
                          #With openpose: BODY_25B, BODY_25, BODY_135, COCO, MPII
@@ -46,9 +44,6 @@ pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 mo
 mode = 'balanced' # 'lightweight', 'balanced', 'performance'
 det_frequency = 1 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
                   # Equal to or greater than 1, can be as high as you want in simple uncrowded cases. Much faster, but might be less accurate. 
-device = 'auto' # 'auto', 'CPU', 'CUDA', 'MPS', 'ROCM'
-backend = 'auto' # 'auto', 'openvino', 'onnxruntime', 'opencv'
-
 display_detection = true
 overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 save_video = 'to_video' # 'to_video' or 'to_images', 'none', or ['to_video', 'to_images']

Pose2Sim/Demo_SinglePerson/Config.toml

@@ -32,11 +32,9 @@ frame_range = [] # For example [10,300], or [] for all frames.
 exclude_from_batch = [] # List of trials to be excluded from batch analysis, ['<participant_dir/trial_dir>', 'etc'].
 # e.g. ['S00_P00_Participant/S00_P00_T00_StaticTrial', 'S00_P00_Participant/S00_P00_T01_BalancingTrial']
 
-
 [pose]
 vid_img_extension = 'mp4' # any video or image extension
-
-pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 model), Whole_body (COCO_133: body + feet + hands), Body (COCO_17)
+pose_model = 'HALPE_26'  #With RTMLib: HALPE_26 (body and feet, default), COCO_133 (body, feet, hands), COCO_17 (body)
                          # /!\ Only RTMPose is natively embeded in Pose2Sim. For all other pose estimation methods, you will have to run them yourself, and then refer to the documentation to convert the output files if needed
                          #With MMPose: HALPE_26, COCO_133, COCO_17, CUSTOM. See CUSTOM example at the end of the file
                          #With openpose: BODY_25B, BODY_25, BODY_135, COCO, MPII
@@ -46,9 +44,6 @@ pose_model = 'Body_with_feet'  #With RTMLib: Body_with_feet (default HALPE_26 mo
 mode = 'balanced' # 'lightweight', 'balanced', 'performance'
 det_frequency = 1 # Run person detection only every N frames, and inbetween track previously detected bounding boxes (keypoint detection is still run on all frames). 
                   # Equal to or greater than 1, can be as high as you want in simple uncrowded cases. Much faster, but might be less accurate. 
-device = 'auto' # 'auto', 'CPU', 'CUDA', 'MPS', 'ROCM'
-backend = 'auto' # 'auto', 'openvino', 'onnxruntime', 'opencv'
-
 display_detection = true
 overwrite_pose = false # set to false if you don't want to recalculate pose estimation when it has already been done
 save_video = 'to_video' # 'to_video' or 'to_images', 'none', or ['to_video', 'to_images']
@@ -186,7 +181,7 @@ make_c3d = true # save triangulated data in c3d format in addition to trc
 
 
 [kinematics]
-use_augmentation = false  # true or false (lowercase) # Set to true if you want to use the model with augmented markers
+use_augmentation = true  # true or false (lowercase) # Set to true if you want to use the model with augmented markers
 right_left_symmetry = true # true or false (lowercase) # Set to false only if you have good reasons to think the participant is not symmetrical (e.g. prosthetic limb)
 
 remove_individual_scaling_setup = true # true or false (lowercase) # If true, the individual scaling setup files are removed to avoid cluttering

Pose2Sim/kinematics.py

@@ -734,16 +734,8 @@ def kinematics_all(config_dict):
     # if single trial
     session_dir = session_dir if 'Config.toml' in os.listdir(session_dir) else os.getcwd()
     use_augmentation = config_dict.get('kinematics').get('use_augmentation')
-    if use_augmentation: 
-        model_name = 'LSTM'
-    else: 
-        model_name = config_dict.get('pose').get('pose_model').upper()
-        if model_name.upper() == 'BODY_WITH_FEET': model_name = 'HALPE_26'
-        elif model_name.upper() == 'WHOLE_BODY': model_name = 'COCO_133'
-        elif model_name.upper() == 'BODY': model_name = 'COCO_17'
-        else:
-            raise ValueError(f"Invalid model_type: {model_name}. Must be 'HALPE_26', 'COCO_133', or 'COCO_17'. Use another network (MMPose, DeepLabCut, OpenPose, AlphaPose, BlazePose...) and convert the output files if you need another model. See documentation.")
-
+    if use_augmentation: model_name = 'LSTM'
+    else: model_name = config_dict.get('pose').get('pose_model').upper()
     right_left_symmetry = config_dict.get('kinematics').get('right_left_symmetry')
     subject_height = config_dict.get('project').get('participant_height')
     subject_mass = config_dict.get('project').get('participant_mass')
@@ -804,13 +796,13 @@ def kinematics_all(config_dict):
     for p, trc_file in enumerate(trc_files):
         logging.info(f"Processing TRC file: {trc_file.resolve()}")
 
-        logging.info("\nScaling...")
+        logging.info("Scaling...")
         perform_scaling(trc_file, kinematics_dir, osim_setup_dir, model_name, right_left_symmetry=right_left_symmetry, subject_height=subject_height[p], subject_mass=subject_mass[p], 
                         remove_scaling_setup=remove_scaling_setup, fastest_frames_to_remove_percent=fastest_frames_to_remove_percent, large_hip_knee_angles=large_hip_knee_angles, trimmed_extrema_percent=trimmed_extrema_percent,close_to_zero_speed_m=close_to_zero_speed_m)
         logging.info(f"\tDone. OpenSim logs saved to {opensim_logs_file.resolve()}.")
         logging.info(f"\tScaled model saved to {(kinematics_dir / (trc_file.stem + '_scaled.osim')).resolve()}")
 
-        logging.info("\nInverse Kinematics...")
+        logging.info("Inverse Kinematics...")
         perform_IK(trc_file, kinematics_dir, osim_setup_dir, model_name, remove_IK_setup=remove_IK_setup)
         logging.info(f"\tDone. OpenSim logs saved to {opensim_logs_file.resolve()}.")
         logging.info(f"\tJoint angle data saved to {(kinematics_dir / (trc_file.stem + '.mot')).resolve()}\n")

Pose2Sim/personAssociation.py

@@ -664,14 +664,7 @@ def associate_all(config_dict):
 
     # selection of tracked keypoint id
     try: # from skeletons.py
-        if pose_model.upper() == 'BODY_WITH_FEET': pose_model = 'HALPE_26'
-        elif pose_model.upper() == 'WHOLE_BODY': pose_model = 'COCO_133'
-        elif pose_model.upper() == 'BODY': pose_model = 'COCO_17'
-        else:
-            raise ValueError(f"Invalid model_type: {pose_model}. Must be 'HALPE_26', 'COCO_133', or 'COCO_17'. Use another network (MMPose, DeepLabCut, OpenPose, AlphaPose, BlazePose...) and convert the output files if you need another model. See documentation.")
-
         model = eval(pose_model)
-
     except:
         try: # from Config.toml
             model = DictImporter().import_(config_dict.get('pose').get(pose_model))