Multimodal fusion to obtain scores by audio and video FI V2
Import required packages
[2]:
from oceanai.modules.lab.build import Run
Build
[3]:
_b5 = Run(
lang = 'en', # Inference language
color_simple = '#333', # Plain text color (hexadecimal code)
color_info = '#1776D2', # The color of the text containing the information (hexadecimal code)
color_err = '#FF0000', # Error text color (hexadecimal code)
color_true = '#008001', # Text color containing positive information (hexadecimal code)
bold_text = True, # Bold text
num_to_df_display = 30, # Number of rows to display in tables
text_runtime = 'Runtime', # Runtime text
metadata = True # Displaying information about library
)
[2023-12-14 22:46:31] OCEANAI - personal traits: Authors: Elena Ryumina [ryumina_ev@mail.ru] Dmitry Ryumin [dl_03.03.1991@mail.ru] Alexey Karpov [karpov@iias.spb.su] Maintainers: Elena Ryumina [ryumina_ev@mail.ru] Dmitry Ryumin [dl_03.03.1991@mail.ru] Version: 1.0.0a16 License: BSD License
Getting and displaying versions of installed libraries
_b5.df_pkgs_- DataFrame with versions of installed libraries
[4]:
_b5.libs_vers(runtime = True, run = True)
| Package | Version | |
|---|---|---|
| 1 | TensorFlow | 2.15.0 |
| 2 | Keras | 2.15.0 |
| 3 | OpenCV | 4.8.1 |
| 4 | MediaPipe | 0.9.0 |
| 5 | NumPy | 1.26.2 |
| 6 | SciPy | 1.11.4 |
| 7 | Pandas | 2.1.3 |
| 8 | Scikit-learn | 1.3.2 |
| 9 | OpenSmile | 2.5.0 |
| 10 | Librosa | 0.10.1 |
| 11 | AudioRead | 3.0.1 |
| 12 | IPython | 8.18.1 |
| 13 | PyMediaInfo | 6.1.0 |
| 14 | Requests | 2.31.0 |
| 15 | JupyterLab | 4.0.9 |
| 16 | LIWC | 0.5.0 |
| 17 | Transformers | 4.36.0 |
| 18 | Sentencepiece | 0.1.99 |
| 19 | Torch | 2.0.1+cpu |
| 20 | Torchaudio | 2.0.2+cpu |
— Runtime: 0.006 sec. —
Analysing audio information (forming model and loading model weights)
Formation of the neural network architecture of the model for obtaining scores by hand-crafted features (audio modality)
_b5.audio_model_hc_- Neural network model tf.keras.Model for obtaining scores by hand-crafted features
[5]:
res_load_audio_model_hc = _b5.load_audio_model_hc(
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:31] Formation of the neural network architecture of the model for obtaining scores by hand-crafted features (audio modality) …
— Runtime: 0.322 sec. —
Downloading the weights of the neural network model to obtain scores by hand-crafted features (audio modality)
_b5.audio_model_hc_- Neural network model tf.keras.Model for obtaining scores by hand-crafted features
[6]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url = _b5.weights_for_big5_['audio']['fi']['hc']['sberdisk']
res_load_audio_model_weights_hc = _b5.load_audio_model_weights_hc(
url = url, # Full path to the file with weights of the neural network model
force_reload = True, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:31] Downloading the weights of the neural network model to obtain scores by hand-crafted features (audio modality) …
[2023-12-14 22:46:32] File download “weights_2022-05-05_11-27-55.h5” (100.0%) …
— Runtime: 0.277 sec. —
Formation of the neural network architecture of the model for obtaining scores by deep features (audio modality)
_b5.audio_model_nn_- Neural network model tf.keras.Model for obtaining scores by deep features
[7]:
res_load_audio_model_nn = _b5.load_audio_model_nn(
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:32] Formation of a neural network architecture for obtaining scores by deep features (audio modality) …
— Runtime: 0.244 sec. —
Downloading the weights of the neural network model to obtain scores for deep features (audio modality)
_b5.audio_model_nn_- Neural network model tf.keras.Model for obtaining scores by deep features
[8]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url = _b5.weights_for_big5_['audio']['fi']['nn']['sberdisk']
res_load_audio_model_weights_nn = _b5.load_audio_model_weights_nn(
url = url, # Full path to the file with weights of the neural network model
force_reload = False, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:32] Downloading the weights of the neural network model to obtain scores for deep features (audio modality) …
[2023-12-14 22:46:32] File download “weights_2022-05-03_07-46-14.h5”
— Runtime: 0.389 sec. —
Analysing video information (forming model and loading model weights)
Formation of the neural network architecture of the model for obtaining scores by hand-crafted features (video modality)
_b5.video_model_hc_- Neural network model tf.keras.Model for obtaining scores by hand-crafted features
[9]:
res_load_video_model_hc = _b5.load_video_model_hc(
lang = 'en', # Language selection for models trained on First Impressions V2'en' and models trained on for MuPTA 'ru'
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:32] Formation of the neural network architecture of the model for obtaining scores by hand-crafted features (video modality) …
— Runtime: 0.257 sec. —
Downloading the weights of the neural network model to obtain scores by hand-crafted features (video modality)
_b5.video_model_hc_- Neural network model tf.keras.Model for obtaining scores by hand-crafted features
[10]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url = _b5.weights_for_big5_['video']['fi']['hc']['sberdisk']
res_load_video_model_weights_hc = _b5.load_video_model_weights_hc(
url = url, # Full path to the file with weights of the neural network model
force_reload = True, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:32] Downloading the weights of the neural network model to obtain scores by hand-crafted features (video modality) …
[2023-12-14 22:46:33] File download “weights_2022-08-27_18-53-35.h5” (100.0%) …
— Runtime: 0.226 sec. —
Formation of neural network architecture for obtaining deep features
_b5.video_model_deep_fe_- Neural network model tf.keras.Model for obtaining deep features
[11]:
res_load_video_model_deep_fe = _b5.load_video_model_deep_fe(
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:34] Formation of neural network architecture for obtaining deep features (video modality) …
— Runtime: 0.783 sec. —
Downloading weights of a neural network model to obtain deep features
_b5.video_model_deep_fe_- Neural network model tf.keras.Model for obtaining deep features
[12]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url = _b5.weights_for_big5_['video']['fi']['fe']['sberdisk']
res_load_video_model_weights_deep_fe = _b5.load_video_model_weights_deep_fe(
url = url, # Full path to the file with weights of the neural network model
force_reload = True, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:35] Downloading weights of a neural network model to obtain deep features (video modality) …
[2023-12-14 22:46:40] File download “weights_2022-11-01_12-27-07.h5” (100.0%) …
— Runtime: 4.311 sec. —
Formation of the neural network architecture of the model for obtaining scores by deep features (video modality)
_b5.video_model_nn_- Neural network model tf.keras.Model for obtaining scores by deep features
[13]:
res_load_video_model_nn = _b5.load_video_model_nn(
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:40] Formation of a neural network architecture for obtaining scores by deep features (video modality) …
— Runtime: 0.724 sec. —
Downloading the weights of the neural network model to obtain scores for deep features (video modality)
_b5.video_model_nn_- Neural network model tf.keras.Model for obtaining scores by deep features
[14]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url = _b5.weights_for_big5_['video']['fi']['nn']['sberdisk']
res_load_video_model_weights_nn = _b5.load_video_model_weights_nn(
url = url, # Full path to the file with weights of the neural network model
force_reload = False, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:40] Downloading the weights of the neural network model to obtain scores by deep features (video modality) …
[2023-12-14 22:46:42] File downloading “weights_2022-03-22_16-31-48.h5”
— Runtime: 1.355 sec. —
Analysing multimodal information (forming model, loading model weights, obtaining personality traits scores)
Formation of neural network architectures of models for obtaining the personality traits scores (multimodal fusion)
_b5.av_models_b5_- Neural network models tf.keras.Model for obtaining the personality traits scores
[15]:
res_load_av_models_b5 = _b5.load_av_models_b5(
show_summary = False, # Displaying the formed neural network architecture of the model
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:42] Formation of neural network architectures of models for obtaining the personality traits scores (multimodal fusion) …
— Runtime: 0.048 sec. —
ЗDownloading the weights of neural network models to obtain the personality traits scores (multimodal fusion)
_b5.av_models_b5_- Neural network models tf.keras.Model for obtaining the personality traits scores
[16]:
# Core settings
_b5.path_to_save_ = './models' # Directory to save the file
_b5.chunk_size_ = 2000000 # File download size from network in 1 step
url_openness = _b5.weights_for_big5_['av']['fi']['b5']['openness']['sberdisk']
url_conscientiousness = _b5.weights_for_big5_['av']['fi']['b5']['conscientiousness']['sberdisk']
url_extraversion = _b5.weights_for_big5_['av']['fi']['b5']['extraversion']['sberdisk']
url_agreeableness = _b5.weights_for_big5_['av']['fi']['b5']['agreeableness']['sberdisk']
url_non_neuroticism = _b5.weights_for_big5_['av']['fi']['b5']['non_neuroticism']['sberdisk']
res_load_av_models_weights_b5 = _b5.load_av_models_weights_b5(
url_openness = url_openness, # Openness
url_conscientiousness = url_conscientiousness, # Conscientiousness
url_extraversion = url_extraversion, # Extraversion
url_agreeableness = url_agreeableness, # Agreeableness
url_neuroticism = url_neuroticism, # Non-Neuroticism
force_reload = True, # Forced download of a file with weights of a neural network model from the network
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-14 22:46:47] Downloading the weights of neural network models to obtain the personality traits scores (multimodal fusion) …
[2023-12-14 22:46:47] File download “weights_2022-08-28_11-14-35.h5” Openness
[2023-12-14 22:46:47] File download “weights_2022-08-28_11-08-10.h5” Conscientiousness
[2023-12-14 22:46:47] File download “weights_2022-08-28_11-17-57.h5” Extraversion
[2023-12-14 22:46:47] File download “weights_2022-08-28_11-25-11.h5” Agreeableness
[2023-12-14 22:46:47] File download “weights_2022-06-14_21-44-09.h5” Non-Neuroticism
— Runtime: 0.785 sec. —
Getting scores (multimodal fusion)
_b5.df_files_- DataFrame with data
_b5.df_accuracy_- DataFrame with accuracy
[17]:
# Core settings
_b5.path_to_dataset_ = 'E:/Databases/FirstImpressionsV2/test' # Dataset directory
# Directories not included in the selection
_b5.ignore_dirs_ = []
# Key names for DataFrame dataset
_b5.keys_dataset_ = ['Path', 'Openness', 'Conscientiousness', 'Extraversion', 'Agreeableness', 'Non-Neuroticism']
_b5.ext_ = ['.mp4'] # Search file extensions
# Full path to the file containing the ground truth scores for the accuracy calculation
url_accuracy = _b5.true_traits_['fi']['sberdisk']
_b5.get_av_union_predictions(
depth = 2, # Hierarchy depth for receiving audio and video data
recursive = False, # Recursive data search
sr = 44100, # Sampling frequency
window_audio = 2, # Audio segment window size (in seconds)
step_audio = 1, # Audio segment window shift step (in seconds)
reduction_fps = 5, # Frame rate reduction
window_video = 10, # Video segment window size (in seconds)
step_video = 5, # Video segment window shift step (in seconds)
lang = 'en', # Language selection for models trained on First Impressions V2'en' and models trained on for MuPTA 'ru'
accuracy = True, # Accuracy
url_accuracy = url_accuracy,
logs = True, # If necessary, generate a LOG file
out = True, # Display
runtime = True, # Runtime count
run = True # Run blocking
)
[2023-12-15 01:11:04] Getting scores and accuracy calculation (multimodal fusion) …
2000 from 2000 (100.0%) … test80_25_Q4wOgixh7E.004.mp4 …
| Path | Openness | Conscientiousness | Extraversion | Agreeableness | Non-Neuroticism | |
|---|---|---|---|---|---|---|
| ID | ||||||
| 1 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.554249 | 0.506548 | 0.440194 | 0.540235 | 0.48605 |
| 2 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.558823 | 0.442357 | 0.50397 | 0.558767 | 0.521587 |
| 3 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.477549 | 0.568616 | 0.333939 | 0.491873 | 0.458966 |
| 4 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.662656 | 0.621852 | 0.58996 | 0.599038 | 0.636035 |
| 5 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.645876 | 0.532378 | 0.551939 | 0.589174 | 0.552269 |
| 6 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.67497 | 0.666972 | 0.617604 | 0.610567 | 0.641452 |
| 7 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.39908 | 0.397298 | 0.335823 | 0.497966 | 0.39729 |
| 8 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.577705 | 0.597157 | 0.498064 | 0.640584 | 0.600152 |
| 9 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.543675 | 0.451197 | 0.449555 | 0.482371 | 0.415256 |
| 10 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.54876 | 0.51097 | 0.433856 | 0.579709 | 0.536171 |
| 11 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.546634 | 0.398485 | 0.443701 | 0.518107 | 0.492343 |
| 12 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.459302 | 0.427114 | 0.315686 | 0.495817 | 0.457954 |
| 13 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.309097 | 0.317028 | 0.218514 | 0.372315 | 0.241697 |
| 14 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.643403 | 0.509414 | 0.483608 | 0.503154 | 0.550979 |
| 15 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.65016 | 0.840148 | 0.535299 | 0.710939 | 0.743357 |
| 16 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.598313 | 0.520505 | 0.450767 | 0.486345 | 0.561532 |
| 17 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.571537 | 0.673989 | 0.472203 | 0.615608 | 0.621064 |
| 18 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.552433 | 0.568787 | 0.457108 | 0.613188 | 0.570902 |
| 19 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.658695 | 0.625194 | 0.634877 | 0.612277 | 0.626052 |
| 20 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.660076 | 0.544358 | 0.64178 | 0.604572 | 0.628259 |
| 21 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.543881 | 0.477881 | 0.407731 | 0.555772 | 0.499664 |
| 22 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.537325 | 0.46375 | 0.419255 | 0.499785 | 0.455146 |
| 23 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.464761 | 0.434816 | 0.346836 | 0.428429 | 0.358087 |
| 24 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.633951 | 0.63333 | 0.584644 | 0.615227 | 0.608006 |
| 25 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.4517 | 0.574346 | 0.350136 | 0.526873 | 0.468283 |
| 26 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.602848 | 0.592382 | 0.494679 | 0.539232 | 0.505865 |
| 27 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.586638 | 0.521421 | 0.485391 | 0.530296 | 0.535499 |
| 28 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.689552 | 0.643902 | 0.695799 | 0.646209 | 0.686243 |
| 29 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.583505 | 0.564313 | 0.502263 | 0.554502 | 0.539899 |
| 30 | E:\Databases\FirstImpressionsV2\test\test80_01... | 0.642695 | 0.588222 | 0.617706 | 0.615312 | 0.626649 |
[2023-12-15 01:11:04] Trait-wise accuracy …
| Openness | Conscientiousness | Extraversion | Agreeableness | Non-Neuroticism | Mean | |
|---|---|---|---|---|---|---|
| Metrics | ||||||
| MAE | 0.0845 | 0.0802 | 0.0793 | 0.0858 | 0.0847 | 0.0829 |
| Accuracy | 0.9155 | 0.9198 | 0.9207 | 0.9142 | 0.9153 | 0.9171 |
[2023-12-15 01:11:04] Mean absolute error: 0.0829, Accuracy: 0.9171 …
Log files saved successfully …
— Runtime: 8654.754 sec. —
[17]:
True