update in controller

Dockerfile

@@ -30,4 +30,4 @@ RUN pip install --upgrade pip
 RUN pip install -r requirements.txt
 RUN echo "Requirement txt done!"
 
-CMD [ "python", "./process_dir.py" ]
+CMD [ "python", "./process_data.py" ]

dbm_lib/controller/process_feature.py

@@ -51,6 +51,7 @@ def process_acoustic(video_uri, out_dir, dbm_group, r_config):
     if dbm_group != None and len(dbm_group)>0 and 'acoustic' not in dbm_group:
         return
     
+    logger.info('Processing acoustic variables from data in {}'.format(video_uri))
     logger.info('processing audio intensity....')
     intensity.run_intensity(video_uri, out_dir, r_config)
     
@@ -91,6 +92,7 @@ def process_facial(video_uri, out_dir, dbm_group, r_config):
     if dbm_group != None and len(dbm_group)>0 and 'facial' not in dbm_group:
         return
     
+    logger.info('Processing facial variables from data in {}'.format(video_uri))
     logger.info('processing facial asymmetry....')
     face_asymmetry.run_face_asymmetry(video_uri, out_dir, r_config)
     
@@ -114,6 +116,7 @@ def process_movement(video_uri, out_dir, dbm_group, r_config, dlib_model):
     if dbm_group != None and len(dbm_group)>0 and 'movement' not in dbm_group:
         return
     
+    logger.info('Processing movement variables from data in {}'.format(video_uri))
     logger.info('processing head movement....')
     head_motion.run_head_movement(video_uri, out_dir, r_config)
     

dbm_lib/dbm_features/derived_features/derive.py

@@ -21,20 +21,19 @@ def dict_to_df(feature_dict, file):
     final_dict = {k: v for d in feature_dict for k, v in d.items()}
 
     feature_df = pd.DataFrame([final_dict])
-    feature_df['dbm_master_url'] = file
+    feature_df['Filename'] = file
     
     return feature_df
 
-def save_derive_output(df_list, feature, out_loc):
+def save_derive_output(df_list, out_loc):
     """
     Saving derive variable output
     """
+    try:
         if len(df_list)>0:
-        logger.info("Saving derived variable output for {}".format(feature))
+            df = df_list[0]
-        
-        df = pd.concat(df_list, ignore_index=True)
-        feature_dir = 'derive_' + feature
             
+            feature_dir = 'derived_output'
             now = datetime.now()
             dt_string = now.strftime("%d_%m_%YT%H:%M:%S")
 
@@ -45,6 +44,9 @@ def save_derive_output(df_list, feature, out_loc):
                 os.makedirs(out_dir)
             df.to_csv(file_name, index=False)
             
+    except Exception as e:
+        logger.error('Failed to save derived variable csv')
+
 def feature_output(df_fea, exp_var, cal_type):
     """
         Computing mean value of dataframe columns
@@ -64,7 +66,7 @@ def feature_output(df_fea, exp_var, cal_type):
                 exp_val = df_.std(axis = 0, skipna = True)
 
             elif cal_type == 'count':#use case for eye blink
-                exp_var = 'blink_count'
+                exp_var = 'blink'
                 exp_val = (len(df_)/df_[0])*60
 
             elif cal_type == 'pct':
@@ -110,11 +112,12 @@ def compute_feature(raw_df, var_cols, d_cfg_Obj, r_cfg_Obj):
 
     return feature_dict
         
-def calc_derive(input_file, input_dir, output_dir, r_cfg_Obj, d_cfg_Obj, feature):
+def calc_derive(input_file, input_dir, r_cfg_Obj, d_cfg_Obj, feature):
     """
     Calculating derived variable
     """
     df_list = []
+    df = pd.DataFrame()
     for file in input_file:
         
         file_name, _ = os.path.splitext(os.path.basename(file))
@@ -133,7 +136,9 @@ def calc_derive(input_file, input_dir, output_dir, r_cfg_Obj, d_cfg_Obj, feature
                 feature_df = dict_to_df(feature_dict, file)
                 df_list.append(feature_df)
     
-    save_derive_output(df_list, feature, output_dir)
+    if len(df_list)>0:
+        df = pd.concat(df_list, ignore_index=True)
+    return df
 
 def run_derive(input_file, input_dir, output_dir, r_config, d_config):
     """
@@ -144,10 +149,21 @@ def run_derive(input_file, input_dir, output_dir, r_config, d_config):
     feature_group = d_cfg_Obj['FEATURE_GROUP']
     
     #Iterating over feature group
+    df_list = []
     for feature in feature_group:
         try:
             
-            calc_derive(input_file, input_dir, output_dir, r_cfg_Obj, d_cfg_Obj, feature)
+            df_fea = calc_derive(input_file, input_dir, r_cfg_Obj, d_cfg_Obj, feature)
-        except Exception as e:
+            if len(df_fea)>0:
-            logger.error('Failed to process derived variables.')
                 
+                if len(df_list) == 0:
+                    df_list.append(df_fea)
+                else:
+                    result = pd.merge(df_list[0], df_fea, how='outer', on=['Filename'])
+                    df_list = [result]
+                    
+        except Exception as e:
+            logger.error('Failed to process derived variables {}'.format(feature))
+    
+    logger.info("Saving derived variable output...")
+    save_derive_output(df_list, output_dir)

dbm_lib/dbm_features/raw_features/audio/gne.py

@@ -21,7 +21,7 @@ logger=logging.getLogger()
 
 gne_dir = 'audio/glottal_noise'
 ff_dir =  'audio/pitch'
-csv_ext = '_gne_frame.csv'
+csv_ext = '_gne.csv'
 
 def gne_ratio(sound):
     """

dbm_lib/dbm_features/raw_features/audio/hnr.py

@@ -19,7 +19,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 hnr_dir = 'audio/harmonic_noise'
-csv_ext = '_hnr_frame.csv'
+csv_ext = '_hnr.csv'
 error_txt = 'error: length less than 0.064'
 
 def hnr_ratio(filepath):

dbm_lib/dbm_features/raw_features/audio/pause_segment.py

@@ -21,7 +21,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 pause_seg_dir = 'audio/pause_segment'
-csv_ext = '_pause_segment.csv'
+csv_ext = '_pausechar.csv'
 
 def get_timing_cues(seg_starts_sec, seg_ends_sec, r_config):
     """

dbm_lib/dbm_features/raw_features/audio/voice_frame_score.py

@@ -18,7 +18,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 vfs_dir = 'audio/voice_frame_score'
-csv_ext = '_vfs.csv'
+csv_ext = '_voiceprev.csv'
 error_txt = 'error: length less than 0.064'
 
 def audio_pitch_frame(pitch):

dbm_lib/dbm_features/raw_features/movement/eye_blink.py

@@ -26,7 +26,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 movement_expr_dir = 'movement/eye_blink'
-csv_ext = '_eye_blink.csv'
+csv_ext = '_eyeblinks.csv'
 
 def eye_aspect_ratio(eye):
     """

dbm_lib/dbm_features/raw_features/movement/head_motion.py

@@ -19,8 +19,8 @@ logger=logging.getLogger()
 
 h_mov_dir = 'movement/head_movement'
 h_pose_dir = 'movement/head_pose'
-h_mov_ext = '_head_movement.csv'
+h_mov_ext = '_headmov.csv'
-h_pose_ext = '_head_pose.csv'
+h_pose_ext = '_headpose.csv'
 
 def head_pose_dist(of_results):
     """

dbm_lib/dbm_features/raw_features/video/face_asymmetry.py

@@ -26,7 +26,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 face_asym_dir = 'video/face_asymmetry'
-csv_ext = '_face_asymmetry.csv'
+csv_ext = '_facasym.csv'
 
 cv2_color_purple = (254,19,188)
 color_blue = (0,0,1.0)

dbm_lib/dbm_features/raw_features/video/face_au.py

@@ -20,7 +20,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 face_au_dir = 'video/face_au'
-csv_ext = '_face_au.csv'
+csv_ext = '_facau.csv'
 
 
 def extract_col_nm_au(cols):

dbm_lib/dbm_features/raw_features/video/face_emotion_expressivity.py

@@ -20,7 +20,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 face_expr_dir = 'video/face_expressivity'
-csv_ext = '_face_expressivity.csv'
+csv_ext = '_facemo.csv'
 
 #Openface feature extraction
 def of_feature(df_of, cfr, f_cfg):

dbm_lib/dbm_features/raw_features/video/face_landmark.py

@@ -20,7 +20,7 @@ logging.basicConfig(level=logging.INFO)
 logger=logging.getLogger()
 
 face_lmk_dir = 'video/face_landmark'
-csv_ext = '_face_landmark.csv'
+csv_ext = '_faclmk.csv'
 
 def extract_col_nm_lmk(cols):
     """

process_data.py

@@ -0,0 +1,180 @@
+"""
+file_name: process_file
+project_name: DBM
+created: 2020-20-07
+"""
+
+from dbm_lib.config import config_reader, config_raw_feature, config_derive_feature
+from dbm_lib.controller import process_feature as pf
+from dbm_lib.dbm_features.raw_features.video import open_face_process as of
+from dbm_lib.dbm_features.derived_features import derive as der
+
+import pandas as pd
+import os
+import argparse
+import logging
+import glob
+import time
+
+logging.basicConfig(level=logging.INFO)
+logger=logging.getLogger()
+
+OPENFACE_PATH = 'pkg/OpenFace/build/bin/FeatureExtraction'
+DLIB_SHAPE_MODEL = 'pkg/shape_detector/shape_predictor_68_face_landmarks.dat'
+
+def common_video(video_file, args, r_config):
+    """
+    Common function for video feature processing
+    Args:
+        video_file: video file path
+        args: user supplied arguments
+        r_config: raw feature config object
+    """
+    pf.audio_to_wav(video_file)
+    of.process_open_face(video_file, os.path.dirname(video_file), args.output_raw_path, OPENFACE_PATH, args.dbm_group)
+
+    pf.process_facial(video_file, args.output_raw_path, args.dbm_group, r_config)
+    pf.process_acoustic(video_file, args.output_raw_path, args.dbm_group, r_config)
+    pf.remove_file(video_file)
+    pf.process_movement(video_file, args.output_raw_path, args.dbm_group, r_config, DLIB_SHAPE_MODEL)
+
+def process_raw_video_file(args, s_config, r_config):
+    """
+    Processing video file
+    Args:
+        args: user supplied arguments
+        s_config: service config object
+        r_config: raw feature config object
+    """
+    try:
+        if args.output_raw_path != None:
+            video_file = glob.glob(args.input_path)
+
+            if len(video_file)>0:
+                logger.info('Calculating raw variables...')
+                common_video(video_file[0], args, r_config)
+
+            else:
+                logger.info('Enter correct video(*.mp4) file path.')
+                
+    except Exception as e:
+        logger.error('Failed to process mp4 file.')
+        pf.remove_file(video_file[0])
+
+def process_raw_audio_file(args, s_config, r_config):
+    """
+    Processing audio file
+    Args:
+        args: user supplied arguments
+        s_config: service config object
+        r_config: raw feature config object
+    """
+    try:
+        if args.output_raw_path != None:
+            audio_file = glob.glob(args.input_path)
+
+            if len(audio_file)>0:
+                logger.info('Calculating raw variables...')
+                pf.process_acoustic(audio_file[0], args.output_raw_path, args.dbm_group, r_config)
+                
+            else:
+                logger.info('Enter correct audio(*.wav) file path.')
+    except Exception as e:
+        logger.error('Failed to process wav file.')
+    
+def process_raw_video_dir(args, s_config, r_config):
+    """
+    Processing video file
+    Args:
+        args: user supplied arguments
+        s_config: service config object
+        r_config: raw feature config object
+    """
+    if args.output_raw_path != None:
+        vid_loc = glob.glob(args.input_path + '/*.mp4')
+        
+        if len(vid_loc) == 0:
+            logger.info('Directory does not have any MP4 files.')
+            return
+        
+        logger.info('Calculating mp4 raw variables...')
+        for vid_file in vid_loc:
+            try:
+                
+                common_video(vid_file, args, r_config)  
+            except Exception as e:
+                logger.error('Failed to process mp4 file.')
+                pf.remove_file(vid_file)
+
+def process_raw_audio_dir(args, s_config, r_config):
+    """
+    Processing audio file
+    Args:
+        args: user supplied arguments
+        s_config: service config object
+        r_config: raw feature config object
+    """
+    if args.output_raw_path != None:
+        audio_loc = glob.glob(args.input_path + '/*.wav')
+        
+        if len(audio_loc) == 0:
+            logger.info('Directory does not have any WAV files.')
+            return
+        
+        logger.info('Calculating wav raw variables...')
+        for audio in audio_loc:
+            try:
+                
+                pf.process_acoustic(audio, args.output_raw_path, args.dbm_group, r_config)
+            except Exception as e:
+                logger.error('Failed to process wav file.')
+        
+def process_derive(args, r_config, d_config, input_type):
+    """
+    Processing dbm derived variables
+    """
+    if input_type == 'file':
+        input_file = glob.glob(args.input_path)
+    else:
+        input_file = glob.glob(args.input_path + '/*')
+    
+    logger.info('Calculating derived variables...')
+    feature_df = der.run_derive(input_file, args.output_raw_path, args.output_derived_path, r_config, d_config)
+    
+if __name__=="__main__":
+    start_time = time.time()
+    parser = argparse.ArgumentParser(description="Process video/audio......")
+    
+    parser.add_argument("--input_path", help="path to the input files", required=True)
+    
+    parser.add_argument("--output_raw_path", help="path to the raw variable output", required=True)
+    parser.add_argument("--output_derived_path", help="path to the derived variable output")
+    parser.add_argument("--dbm_group", help="list of feature groups", nargs='+')
+    
+    args = parser.parse_args()
+    s_config = config_reader.ConfigReader()
+    r_config = config_raw_feature.ConfigRawReader()
+    d_config = config_derive_feature.ConfigDeriveReader()
+    
+    _, file_ext = os.path.splitext(os.path.basename(args.input_path))
+    
+    if file_ext:
+        input_type = 'file'
+        if file_ext.lower() == '.mp4':
+            process_raw_video_file(args, s_config, r_config)
+
+        elif file_ext.lower() == '.wav':
+            process_raw_audio_file(args, s_config, r_config)
+
+        else:
+            logger.error('No WAV or MP4 files detected in input path')
+    else:
+        input_type = 'dir'
+        process_raw_video_dir(args, s_config, r_config)
+        process_raw_audio_dir(args, s_config, r_config)
+        
+    if args.output_derived_path != None:
+        process_derive(args, r_config, d_config, input_type)
+    
+    exec_time = time.time() - start_time
+    logger.info('Done! Processing time: {} seconds'.format(exec_time))

process_dir.py

@@ -1,107 +0,0 @@
-"""
-file_name: process_dir
-project_name: DBM
-created: 2020-20-07
-"""
-
-from dbm_lib.config import config_reader, config_raw_feature, config_derive_feature
-from dbm_lib.controller import process_feature as pf
-from dbm_lib.dbm_features.raw_features.video import open_face_process as of
-from dbm_lib.dbm_features.derived_features import derive as der
-
-import pandas as pd
-import os
-import argparse
-import logging
-import glob
-
-logging.basicConfig(level=logging.INFO)
-logger=logging.getLogger()
-
-OPENFACE_PATH = 'pkg/OpenFace/build/bin/FeatureExtraction'
-DLIB_SHAPE_MODEL = 'pkg/shape_detector/shape_predictor_68_face_landmarks.dat'
-
-def process_raw_video(args, s_config, r_config):
-    """
-    Processing video file
-    Args:
-        args: user supplied arguments
-        s_config: service config object
-        r_config: raw feature config object
-    """
-    if args.output_raw_path != None:
-        vid_loc = glob.glob(args.input_path + '/*.mp4')
-        
-        if len(vid_loc) == 0:
-            logger.info('Directory does not have any mp4 files.')
-            return
-        
-        logger.info('Processing mp4 files.')
-        for vid_file in vid_loc:
-            try:
-                
-                pf.audio_to_wav(vid_file)
-                of.process_open_face(vid_file, os.path.dirname(vid_file), args.output_raw_path, OPENFACE_PATH, args.dbm_group)
-                
-                pf.process_facial(vid_file, args.output_raw_path, args.dbm_group, r_config)
-                pf.process_acoustic(vid_file, args.output_raw_path, args.dbm_group, r_config)
-                pf.remove_file(vid_file)
-                pf.process_movement(vid_file, args.output_raw_path, args.dbm_group, r_config, DLIB_SHAPE_MODEL)
-                
-            except Exception as e:
-                logger.error('Failed to process mp4 file.')
-                pf.remove_file(vid_file)
-
-def process_raw_audio(args, s_config, r_config):
-    """
-    Processing audio file
-    Args:
-        args: user supplied arguments
-        s_config: service config object
-        r_config: raw feature config object
-    """
-    if args.output_raw_path != None:
-        audio_loc = glob.glob(args.input_path + '/*.wav')
-        
-        if len(audio_loc) == 0:
-            logger.info('Directory does not have any wav files.')
-            return
-        
-        for audio in audio_loc:
-            try:
-                
-                logger.info('Processing wav files.')
-                pf.process_acoustic(audio, args.output_raw_path, args.dbm_group, r_config)
-                
-            except Exception as e:
-                logger.error('Failed to process wav file.')
-        
-def process_derive(args, r_config, d_config):
-    """
-    Processing dbm derived variables
-    Args:
-        s_config: service config object
-    """
-    input_file = glob.glob(args.input_path + '/*')
-    feature_df = der.run_derive(input_file, args.output_raw_path, args.output_derived_path, r_config, d_config)
-
-if __name__=="__main__":
-    parser = argparse.ArgumentParser(description="Process video/audio......")
-    
-    parser.add_argument("--input_path", help="path to the input directory", required=True)
-    
-    parser.add_argument("--output_raw_path", help="dir path to the raw variable output", required=True)
-    parser.add_argument("--output_derived_path", help="dir path to the derived variable output")
-    parser.add_argument("--dbm_group", help="list of feature groups", nargs='+')
-    
-    args = parser.parse_args() 
-    s_config = config_reader.ConfigReader()
-    r_config = config_raw_feature.ConfigRawReader()
-    d_config = config_derive_feature.ConfigDeriveReader()
-    
-    process_raw_video(args, s_config, r_config)
-    process_raw_audio(args, s_config, r_config)
-    
-    if args.output_derived_path != None:
-        process_derive(args, r_config, d_config)
-    

process_file.py

@@ -1,107 +0,0 @@
-"""
-file_name: process_file
-project_name: DBM
-created: 2020-20-07
-"""
-
-from dbm_lib.config import config_reader, config_raw_feature, config_derive_feature
-from dbm_lib.controller import process_feature as pf
-from dbm_lib.dbm_features.raw_features.video import open_face_process as of
-from dbm_lib.dbm_features.derived_features import derive as der
-
-import pandas as pd
-import os
-import argparse
-import logging
-import glob
-
-logging.basicConfig(level=logging.INFO)
-logger=logging.getLogger()
-
-OPENFACE_PATH = 'pkg/OpenFace/build/bin/FeatureExtraction'
-DLIB_SHAPE_MODEL = 'pkg/shape_detector/shape_predictor_68_face_landmarks.dat'
-
-def process_raw_video(args, s_config, r_config):
-    """
-    Processing video file
-    Args:
-        args: user supplied arguments
-        s_config: service config object
-        r_config: raw feature config object
-    """
-    try:
-        if args.output_raw_path != None:
-            video_file = glob.glob(args.input_path)
-
-            if len(video_file)>0:
-                pf.audio_to_wav(video_file[0])
-                of.process_open_face(video_file[0], os.path.dirname(video_file[0]), args.output_raw_path, 
-                                     OPENFACE_PATH, args.dbm_group)
-
-                pf.process_facial(video_file[0], args.output_raw_path, args.dbm_group, r_config)
-                pf.process_acoustic(video_file[0], args.output_raw_path, args.dbm_group, r_config)
-                pf.remove_file(video_file[0])
-                pf.process_movement(video_file[0], args.output_raw_path, args.dbm_group, r_config, DLIB_SHAPE_MODEL)
-
-            else:
-                logger.info('Enter correct video(*.mp4) file path.')
-                
-    except Exception as e:
-        logger.error('Failed to process mp4 file.')
-        pf.remove_file(video_file[0])
-
-def process_raw_audio(args, s_config, r_config):
-    """
-    Processing audio file
-    Args:
-        args: user supplied arguments
-        s_config: service config object
-        r_config: raw feature config object
-    """
-    try:
-        if args.output_raw_path != None:
-            audio_file = glob.glob(args.input_path)
-
-            if len(audio_file)>0:
-                pf.process_acoustic(audio_file[0], args.output_raw_path, args.dbm_group, r_config)
-            else:
-                logger.info('Enter correct audio(*.wav) file path.')
-    except Exception as e:
-        logger.error('Failed to process wav file.')
-        
-def process_derive(args, r_config, d_config):
-    """
-    Processing dbm derived variables
-    """
-    input_file = glob.glob(args.input_path)
-    feature_df = der.run_derive(input_file, args.output_raw_path, args.output_derived_path, r_config, d_config)
-    
-
-if __name__=="__main__":
-    parser = argparse.ArgumentParser(description="Process video/audio......")
-    
-    parser.add_argument("--input_path", help="path to the input files", required=True)
-    
-    parser.add_argument("--output_raw_path", help="path to the raw variable output", required=True)
-    parser.add_argument("--output_derived_path", help="path to the derived variable output")
-    parser.add_argument("--dbm_group", help="list of feature groups", nargs='+')
-    
-    args = parser.parse_args()
-    s_config = config_reader.ConfigReader()
-    r_config = config_raw_feature.ConfigRawReader()
-    d_config = config_derive_feature.ConfigDeriveReader()
-    
-    _, file_ext = os.path.splitext(os.path.basename(args.input_path))
-    
-    if file_ext.lower() == '.mp4':
-        process_raw_video(args, s_config, r_config)
-    
-    elif file_ext.lower() == '.wav':
-        process_raw_audio(args, s_config, r_config)
-        
-    else:
-        logger.info('Pipeline accepts mp4 or wav file only.')
-    
-    if args.output_derived_path != None:
-        process_derive(args, r_config, d_config)
-    

resources/features/derived_feature.yml

@@ -5,23 +5,23 @@ derive_feature:
     'ACO_JITTER','ACO_SHIMMER', 'ACO_PAUSE', 'ACO_VFS', 'ACO_MFCC', 'MOV_HM', 'MOV_HP', 'EYE_BLINK']
     
     #Feature group output file extensions
-    FAC_ASYM_LOC: _face_asymmetry
+    FAC_ASYM_LOC: _facasym
-    FAC_AU_LOC: _face_au
+    FAC_AU_LOC: _facau
-    FAC_EXP_LOC: _face_expressivity
+    FAC_EXP_LOC: _facemo
-    FAC_LMK_LOC: _face_landmark
+    FAC_LMK_LOC: _faclmk
     ACO_INT_LOC: _intensity
     ACO_FF_LOC: _pitch
-    ACO_HNR_LOC: _hnr_frame
+    ACO_HNR_LOC: _hnr
-    ACO_GNE_LOC: _gne_frame
+    ACO_GNE_LOC: _gne
     ACO_FM_LOC: _formant
     ACO_JITTER_LOC: _jitter
     ACO_SHIMMER_LOC: _shimmer
-    ACO_PAUSE_LOC: _pause_segment
+    ACO_PAUSE_LOC: _pausechar
-    ACO_VFS_LOC: _vfs
+    ACO_VFS_LOC: _voiceprev
     ACO_MFCC_LOC: _mfcc
-    MOV_HM_LOC: _head_movement
+    MOV_HM_LOC: _headmov
-    MOV_HP_LOC: _head_pose
+    MOV_HP_LOC: _headpose
-    EYE_BLINK_LOC: _eye_blink
+    EYE_BLINK_LOC: _eyeblinks
     
     #Facial category feature group
     FAC_ASYM: ['fac_AsymMaskMouth', 'fac_AsymMaskEyebrow', 'fac_AsymMaskEye', 'fac_AsymMaskCom']