Skip to content

2
2020-101
Commit 765ece4e authored by I.K Seneviratne's avatar I.K Seneviratne
Browse files
Options

Committing the partial implementation of the workflow for lecture student behavior.

parent 033e0193
Hide whitespace changes
Inline Side-by-side
Showing with 432 additions and 161 deletions
FirstApp/api.py
View file @ 765ece4e
...@@ -279,7 +279,7 @@ class LectureActivityProcess(APIView): ...@@ -279,7 +279,7 @@ class LectureActivityProcess(APIView):
def get(self, request): def get(self, request):
video_name = request.query_params.get('lecture_video_name') video_name = request.query_params.get('lecture_video_name')
video_id = request.query_params.get('lecture_video_id') video_id = int(request.query_params.get('lecture_video_id'))
percentages = ar.activity_recognition(video_name) percentages = ar.activity_recognition(video_name)
self.activity(video_id, percentages) self.activity(video_id, percentages)
return Response({"response": True}) return Response({"response": True})
...@@ -288,9 +288,10 @@ class LectureActivityProcess(APIView): ...@@ -288,9 +288,10 @@ class LectureActivityProcess(APIView):
pass pass
def activity(self, lec_video_id, percentages): def activity(self, lec_video_id, percentages):
lec_video = LectureVideo.objects.get(lecture_video_id=lec_video_id) lec_video = LectureVideo.objects.filter(lecture_video_id=lec_video_id)
last_lec_activity = LectureActivity.objects.order_by('lecture_activity_id').last() last_lec_activity = LectureActivity.objects.order_by('lecture_activity_id').last()
lec_video_serializer = LectureVideoSerializer(lec_video, many=True) lec_video_serializer = LectureVideoSerializer(lec_video, many=True)
lec_video_data = lec_video_serializer.data[0]
new_lecture_activity_id = ig.generate_new_id(last_lec_activity.lecture_activity_id) new_lecture_activity_id = ig.generate_new_id(last_lec_activity.lecture_activity_id)
# creating a new lecture activity # creating a new lecture activity
...@@ -303,6 +304,21 @@ class LectureActivityProcess(APIView): ...@@ -303,6 +304,21 @@ class LectureActivityProcess(APIView):
writing_perct=percentages['writing_perct'] writing_perct=percentages['writing_perct']
).save() ).save()
# get the video name
video_name = lec_video_data['video_name']
# then save the frame recognitions to the database
_ = ar.save_frame_recognition(video_name)
# save the time landmarks and frame landmarks
ve.save_time_landmarks(video_name)
frame_landmarks, frame_group_dict = ve.save_frame_landmarks(video_name)
# then save the activity frame groupings
ar.save_frame_groupings(video_name, frame_landmarks, frame_group_dict)
class GetLectureActivityDetections(APIView): class GetLectureActivityDetections(APIView):
...@@ -378,41 +394,8 @@ class GetLectureActivityRecognitionsForFrames(APIView): ...@@ -378,41 +394,8 @@ class GetLectureActivityRecognitionsForFrames(APIView):
else: else:
# retrieve the lecture activity id # perform the action of saving frame recognitions to database
lec_activity = LectureActivity.objects.filter(lecture_video_id__video_name=video_name) frame_detections = ar.save_frame_recognition(video_name)
lec_activity_ser = LectureActivitySerializer(lec_activity, many=True)
lec_activity_data = lec_activity_ser.data[0]
lec_activity_id = lec_activity_data['id']
# create a new lecture activity frame detections id
last_lec_activity_frame_recognitions = LectureActivityFrameRecognitions.objects.order_by('lecture_activity_frame_recognition_id').last()
new_lecture_activity_frame_recognitions_id = "LAFR00001" if (last_lec_activity_frame_recognitions is None) else \
ig.generate_new_id(last_lec_activity_frame_recognitions.lecture_activity_frame_recognition_id)
# calculate the frame detections
frame_detections = ar.get_frame_activity_recognition(video_name)
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_activity_frame_recognition_details = LectureActivityFrameRecognitionDetails()
lec_activity_frame_recognition_details.frame_name = detection['frame_name']
lec_activity_frame_recognition_details.phone_perct = detection['phone_perct']
lec_activity_frame_recognition_details.listen_perct = detection['listening_perct']
lec_activity_frame_recognition_details.note_perct = detection['note_perct']
frame_recognition_details.append(lec_activity_frame_recognition_details)
lec_activity_frame_recognitions = LectureActivityFrameRecognitions()
lec_activity_frame_recognitions.lecture_activity_frame_recognition_id = new_lecture_activity_frame_recognitions_id
lec_activity_frame_recognitions.lecture_activity_id_id = lec_activity_id
lec_activity_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_activity_frame_recognitions.save()
return Response({ return Response({
"response": frame_detections "response": frame_detections
...@@ -490,6 +473,7 @@ class LectureEmotionProcess(APIView): ...@@ -490,6 +473,7 @@ class LectureEmotionProcess(APIView):
def save_emotion_report(self, lec_video_id, percentages): def save_emotion_report(self, lec_video_id, percentages):
lec_video = LectureVideo.objects.get(lecture_video_id=lec_video_id) lec_video = LectureVideo.objects.get(lecture_video_id=lec_video_id)
lec_video_serializer = LectureVideoSerializer(lec_video, many=True) lec_video_serializer = LectureVideoSerializer(lec_video, many=True)
lec_video_data = lec_video_serializer.data[0]
last_lec_emotion = LectureEmotionReport.objects.order_by('lecture_emotion_id').last() last_lec_emotion = LectureEmotionReport.objects.order_by('lecture_emotion_id').last()
new_lecture_emotion_id = ig.generate_new_id(last_lec_emotion.lecture_emotion_id) new_lecture_emotion_id = ig.generate_new_id(last_lec_emotion.lecture_emotion_id)
...@@ -504,6 +488,20 @@ class LectureEmotionProcess(APIView): ...@@ -504,6 +488,20 @@ class LectureEmotionProcess(APIView):
surprise_perct=percentages.surprise_perct surprise_perct=percentages.surprise_perct
).save() ).save()
# get the video name
video_name = lec_video_data['video_name']
# then save the frame recognition details to the database
_ = ed.save_frame_recognitions(video_name)
# retrieve the frame landmarks and frame group dictionary
frame_landmarks, frame_group_dict = ve.getFrameLandmarks(video_name, "Emotion")
# then save emotion frame groupings
ed.save_frame_groupings(video_name, frame_landmarks, frame_group_dict)
# to get a lecture emotion report # to get a lecture emotion report
class GetLectureEmotionReportViewSet(APIView): class GetLectureEmotionReportViewSet(APIView):
...@@ -576,45 +574,8 @@ class GetLectureEmotionRecognitionsForFrames(APIView): ...@@ -576,45 +574,8 @@ class GetLectureEmotionRecognitionsForFrames(APIView):
}) })
else: else:
# save the frame recognitions into the database
# retrieve the lecture emotion id frame_detections = ed.save_frame_recognitions(video_name)
lec_emotion = LectureEmotionReport.objects.filter(lecture_video_id__video_name=video_name)
lec_emotion_ser = LectureEmotionSerializer(lec_emotion, many=True)
lec_emotion_data = lec_emotion_ser.data[0]
lec_emotion_id = lec_emotion_data['id']
# create a new lecture activity frame detections id
last_lec_emotion_frame_recognitions = LectureEmotionFrameRecognitions.objects.order_by(
'lecture_emotion_frame_recognition_id').last()
new_lecture_emotion_frame_recognitions_id = "LEFR00001" if (
last_lec_emotion_frame_recognitions is None) else \
ig.generate_new_id(last_lec_emotion_frame_recognitions.lecture_emotion_frame_recognition_id)
# calculate the frame detections
frame_detections = ed.get_frame_emotion_recognition(video_name)
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_emotion_frame_recognition_details = LectureEmotionFrameRecognitionDetails()
lec_emotion_frame_recognition_details.frame_name = detection['frame_name']
lec_emotion_frame_recognition_details.happy_perct = detection['happy_perct']
lec_emotion_frame_recognition_details.sad_perct = detection['sad_perct']
lec_emotion_frame_recognition_details.angry_perct = detection['angry_perct']
lec_emotion_frame_recognition_details.surprise_perct = detection['surprise_perct']
lec_emotion_frame_recognition_details.neutral_perct = detection['neutral_perct']
frame_recognition_details.append(lec_emotion_frame_recognition_details)
lec_emotion_frame_recognitions = LectureEmotionFrameRecognitions()
lec_emotion_frame_recognitions.lecture_emotion_frame_recognition_id = new_lecture_emotion_frame_recognitions_id
lec_emotion_frame_recognitions.lecture_emotion_id_id = lec_emotion_id
lec_emotion_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_emotion_frame_recognitions.save()
return Response({ return Response({
"response": frame_detections "response": frame_detections
...@@ -725,6 +686,7 @@ class ProcessLectureGazeEstimation(APIView): ...@@ -725,6 +686,7 @@ class ProcessLectureGazeEstimation(APIView):
lec_video = LectureVideo.objects.get(lecture_video_id=lec_video_id) lec_video = LectureVideo.objects.get(lecture_video_id=lec_video_id)
last_lec_gaze = LectureGazeEstimation.objects.order_by('lecture_gaze_id').last() last_lec_gaze = LectureGazeEstimation.objects.order_by('lecture_gaze_id').last()
lec_video_serializer = LectureVideoSerializer(lec_video, many=True) lec_video_serializer = LectureVideoSerializer(lec_video, many=True)
lec_video_data = lec_video_serializer.data[0]
new_lecture_gaze_id = "LG000001" if (last_lec_gaze is None) else ig.generate_new_id( new_lecture_gaze_id = "LG000001" if (last_lec_gaze is None) else ig.generate_new_id(
last_lec_gaze.lecture_gaze_id) last_lec_gaze.lecture_gaze_id)
...@@ -739,6 +701,18 @@ class ProcessLectureGazeEstimation(APIView): ...@@ -739,6 +701,18 @@ class ProcessLectureGazeEstimation(APIView):
looking_front_perct=percentages['head_front_perct'] looking_front_perct=percentages['head_front_perct']
).save() ).save()
# get the video name
video_name = lec_video_data['video_name']
# then save the frame recognitions to the database
_ = hge.save_frame_detections(video_name)
# get the frame landmarks and frame group dictionary
frame_landmarks, frame_group_dict = ve.getFrameLandmarks(video_name, "Gaze")
# then save the gaze frame groupings to the database
hge.save_frame_groupings(video_name, frame_landmarks, frame_group_dict)
# the API to retrieve lecture gaze estimation # the API to retrieve lecture gaze estimation
class GetLectureGazeEstimationViewSet(APIView): class GetLectureGazeEstimationViewSet(APIView):
...@@ -765,7 +739,7 @@ class GetLectureGazeEstimationForFrames(APIView): ...@@ -765,7 +739,7 @@ class GetLectureGazeEstimationForFrames(APIView):
def get(self, request): def get(self, request):
video_name = request.query_params.get('video_name') video_name = request.query_params.get('video_name')
# finding the existence of Lecture activity frame recognition record # finding the existence of Lecture gaze frame recognition record
isExist = LectureGazeFrameRecognitions.objects.filter( isExist = LectureGazeFrameRecognitions.objects.filter(
lecture_gaze_id__lecture_video_id__video_name=video_name).exists() lecture_gaze_id__lecture_video_id__video_name=video_name).exists()
...@@ -784,42 +758,8 @@ class GetLectureGazeEstimationForFrames(APIView): ...@@ -784,42 +758,8 @@ class GetLectureGazeEstimationForFrames(APIView):
else: else:
# retrieve the lecture emotion id # save recognition details into the database
lec_gaze = LectureGazeEstimation.objects.filter(lecture_video_id__video_name=video_name) frame_detections = hge.save_frame_detections(video_name)
lec_gaze_ser = LectureGazeEstimationSerializer(lec_gaze, many=True)
lec_gaze_data = lec_gaze_ser.data[0]
lec_gaze_id = lec_gaze_data['id']
# create a new lecture activity frame detections id
last_lec_gaze_frame_recognitions = LectureGazeFrameRecognitions.objects.order_by(
'lecture_gaze_frame_recognition_id').last()
new_lecture_gaze_frame_recognitions_id = "LGFR00001" if (
last_lec_gaze_frame_recognitions is None) else \
ig.generate_new_id(last_lec_gaze_frame_recognitions.lecture_gaze_frame_recognition_id)
# calculate the frame detections
frame_detections, frame_rate = hge.get_lecture_gaze_esrimation_for_frames(video_name)
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_gaze_frame_recognition_details = LectureGazeFrameRecognitionDetails()
lec_gaze_frame_recognition_details.frame_name = detection['frame_name']
lec_gaze_frame_recognition_details.upright_perct = detection['upright_perct']
lec_gaze_frame_recognition_details.upleft_perct = detection['upleft_perct']
lec_gaze_frame_recognition_details.downright_perct = detection['downright_perct']
lec_gaze_frame_recognition_details.downleft_perct = detection['downleft_perct']
lec_gaze_frame_recognition_details.front_perct = detection['front_perct']
frame_recognition_details.append(lec_gaze_frame_recognition_details)
lec_gaze_frame_recognitions = LectureGazeFrameRecognitions()
lec_gaze_frame_recognitions.lecture_gaze_frame_recognition_id = new_lecture_gaze_frame_recognitions_id
lec_gaze_frame_recognitions.lecture_gaze_id_id = lec_gaze_id
lec_gaze_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_gaze_frame_recognitions.save()
return Response({ return Response({
"response": frame_detections "response": frame_detections
......
FirstApp/emotion_detector.py
View file @ 765ece4e
...@@ -5,11 +5,17 @@ from keras.preprocessing import image ...@@ -5,11 +5,17 @@ from keras.preprocessing import image
import cv2 import cv2
import os import os
import numpy as np import numpy as np
from .MongoModels import *
from . models import VideoMeta from . models import VideoMeta
from . logic import custom_sorter as cs from . logic import custom_sorter as cs
from .logic import id_generator as ig
# emotion recognition method # emotion recognition method
from .serializers import LectureEmotionSerializer
def emotion_recognition(classifier, face_classifier, image): def emotion_recognition(classifier, face_classifier, image):
label = "" label = ""
class_labels = ['Angry', 'Happy', 'Neutral', 'Sad', 'Surprise'] class_labels = ['Angry', 'Happy', 'Neutral', 'Sad', 'Surprise']
...@@ -547,4 +553,83 @@ def emotion_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -547,4 +553,83 @@ def emotion_frame_groupings(video_name, frame_landmarks, frame_group_dict):
# return the dictionary # return the dictionary
return frame_group_dict, emotion_labels return frame_group_dict, emotion_labels
\ No newline at end of file
# this section will handle some database operations
def save_frame_recognitions(video_name):
# retrieve the lecture emotion id
lec_emotion = LectureEmotionReport.objects.filter(lecture_video_id__video_name=video_name)
lec_emotion_ser = LectureEmotionSerializer(lec_emotion, many=True)
lec_emotion_data = lec_emotion_ser.data[0]
lec_emotion_id = lec_emotion_data['id']
# create a new lecture activity frame detections id
last_lec_emotion_frame_recognitions = LectureEmotionFrameRecognitions.objects.order_by(
'lecture_emotion_frame_recognition_id').last()
new_lecture_emotion_frame_recognitions_id = "LEFR00001" if (
last_lec_emotion_frame_recognitions is None) else \
ig.generate_new_id(last_lec_emotion_frame_recognitions.lecture_emotion_frame_recognition_id)
# calculate the frame detections
frame_detections = get_frame_emotion_recognition(video_name)
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_emotion_frame_recognition_details = LectureEmotionFrameRecognitionDetails()
lec_emotion_frame_recognition_details.frame_name = detection['frame_name']
lec_emotion_frame_recognition_details.happy_perct = detection['happy_perct']
lec_emotion_frame_recognition_details.sad_perct = detection['sad_perct']
lec_emotion_frame_recognition_details.angry_perct = detection['angry_perct']
lec_emotion_frame_recognition_details.surprise_perct = detection['surprise_perct']
lec_emotion_frame_recognition_details.neutral_perct = detection['neutral_perct']
frame_recognition_details.append(lec_emotion_frame_recognition_details)
lec_emotion_frame_recognitions = LectureEmotionFrameRecognitions()
lec_emotion_frame_recognitions.lecture_emotion_frame_recognition_id = new_lecture_emotion_frame_recognitions_id
lec_emotion_frame_recognitions.lecture_emotion_id_id = lec_emotion_id
lec_emotion_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_emotion_frame_recognitions.save()
# now return the frame recognitions
return frame_detections
# this method will save the emotion frame groupings to the database
def save_frame_groupings(video_name, frame_landmarks, frame_group_dict):
frame_group_percentages, emotion_labels = emotion_frame_groupings(video_name, frame_landmarks, frame_group_dict)
# save the frame group details into db
last_lec_emotion_frame_grouping = LectureEmotionFrameGroupings.objects.order_by('lecture_emotion_frame_groupings_id').last()
new_lecture_emotion_frame_grouping_id = "LEFG00001" if (last_lec_emotion_frame_grouping is None) else \
ig.generate_new_id(last_lec_emotion_frame_grouping.lecture_emotion_frame_groupings_id)
# retrieve the lecture emotion id
lec_emotion = LectureEmotionReport.objects.filter(lecture_video_id__video_name=video_name)
lec_emotion_ser = LectureEmotionSerializer(lec_emotion, many=True)
lec_emotion_id = lec_emotion_ser.data[0]['id']
# create the frame group details
frame_group_details = []
for key in frame_group_percentages.keys():
# create an object of type 'LectureActivityFrameGroupDetails'
lec_emotion_frame_group_details = LectureEmotionFrameGroupDetails()
lec_emotion_frame_group_details.frame_group = key
lec_emotion_frame_group_details.frame_group_percentages = frame_group_percentages[key]
frame_group_details.append(lec_emotion_frame_group_details)
new_lec_emotion_frame_groupings = LectureEmotionFrameGroupings()
new_lec_emotion_frame_groupings.lecture_emotion_frame_groupings_id = new_lecture_emotion_frame_grouping_id
new_lec_emotion_frame_groupings.lecture_emotion_id_id = lec_emotion_id
new_lec_emotion_frame_groupings.frame_group_details = frame_group_details
# save
new_lec_emotion_frame_groupings.save()
FirstApp/logic/activity_recognition.py
View file @ 765ece4e
...@@ -5,7 +5,10 @@ import numpy as np ...@@ -5,7 +5,10 @@ import numpy as np
import cv2 import cv2
import os import os
import shutil import shutil
from . custom_sorter import * from .custom_sorter import *
from ..MongoModels import *
from ..serializers import *
from . import id_generator as ig
def activity_recognition(video_path): def activity_recognition(video_path):
...@@ -55,30 +58,30 @@ def activity_recognition(video_path): ...@@ -55,30 +58,30 @@ def activity_recognition(video_path):
VIDEO_ACTIVITY_DIR = os.path.join(ACTIVITY_DIR, video_path) VIDEO_ACTIVITY_DIR = os.path.join(ACTIVITY_DIR, video_path)
# creating the directory for the video # creating the directory for the video
if (os.path.isdir(VIDEO_ACTIVITY_DIR)): # if (os.path.isdir(VIDEO_ACTIVITY_DIR)):
shutil.rmtree(VIDEO_ACTIVITY_DIR) # shutil.rmtree(VIDEO_ACTIVITY_DIR)
#
# create the video directory # # create the video directory
os.mkdir(VIDEO_ACTIVITY_DIR) # os.mkdir(VIDEO_ACTIVITY_DIR)
while (frame_count < no_of_frames): while (frame_count < no_of_frames):
ret, image = video.read() ret, image = video.read()
FRAME_DIR = os.path.join(VIDEO_ACTIVITY_DIR, "frame-{}".format(frame_count)) FRAME_DIR = os.path.join(VIDEO_ACTIVITY_DIR, "frame-{}".format(frame_count))
frame_name = "frame-{}.png".format(frame_count) # frame_name = "frame-{}.png".format(frame_count)
#
FRAME_IMG = os.path.join(FRAME_DIR, frame_name) # FRAME_IMG = os.path.join(FRAME_DIR, frame_name)
#
if (os.path.isdir(FRAME_DIR)): # if (os.path.isdir(FRAME_DIR)):
shutil.rmtree(FRAME_DIR) # shutil.rmtree(FRAME_DIR)
# create the new frame directory # create the new frame directory
os.mkdir(FRAME_DIR) # os.mkdir(FRAME_DIR)
image = cv2.resize(image, size) image = cv2.resize(image, size)
detections = person_detection(image, net) detections = person_detection(image, net)
image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
cv2.imwrite(FRAME_IMG, image) # cv2.imwrite(FRAME_IMG, image)
# if there are any person detections # if there are any person detections
if (len(detections) > 0): if (len(detections) > 0):
...@@ -111,22 +114,21 @@ def activity_recognition(video_path): ...@@ -111,22 +114,21 @@ def activity_recognition(video_path):
note_taking_count += 1 note_taking_count += 1
# saving the detection for the particular frame # saving the detection for the particular frame
detection_name = "detection-{}.png".format(detection_count) # detection_name = "detection-{}.png".format(detection_count)
detection_image_path = os.path.join(FRAME_DIR, detection_name) # detection_image_path = os.path.join(FRAME_DIR, detection_name)
#
# converting detected image into grey-scale # # converting detected image into grey-scale
detection = cv2.cvtColor(detection, cv2.COLOR_BGR2GRAY) # detection = cv2.cvtColor(detection, cv2.COLOR_BGR2GRAY)
#
cv2.imwrite(detection_image_path, detection) # cv2.imwrite(detection_image_path, detection)
detection_count += 1 detection_count += 1
frame_count += 1 frame_count += 1
# after extracting the frames, save the changes to static content # after extracting the frames, save the changes to static content
p = os.popen("python manage.py collectstatic", "w") # p = os.popen("python manage.py collectstatic", "w")
p.write("yes") # p.write("yes")
# calculating the percentages for each label # calculating the percentages for each label
phone_perct = float(phone_checking_count / total_detections) * 100 if total_detections > 0 else 0 phone_perct = float(phone_checking_count / total_detections) * 100 if total_detections > 0 else 0
...@@ -140,7 +142,6 @@ def activity_recognition(video_path): ...@@ -140,7 +142,6 @@ def activity_recognition(video_path):
percentages["writing_perct"] = note_perct percentages["writing_perct"] = note_perct
percentages["listening_perct"] = listening_perct percentages["listening_perct"] = listening_perct
return percentages return percentages
...@@ -162,8 +163,6 @@ def person_detection(image, net): ...@@ -162,8 +163,6 @@ def person_detection(image, net):
person_count = 0 person_count = 0
# load the input image and construct an input blob for the image # load the input image and construct an input blob for the image
# by resizing to a fixed 300x300 pixels and then normalizing it # by resizing to a fixed 300x300 pixels and then normalizing it
# (note: normalization is done via the authors of the MobileNet SSD # (note: normalization is done via the authors of the MobileNet SSD
...@@ -212,10 +211,8 @@ def person_detection(image, net): ...@@ -212,10 +211,8 @@ def person_detection(image, net):
# retrieving the extracted frames and detections for a given video # retrieving the extracted frames and detections for a given video
def getExtractedFrames(folder_name): def getExtractedFrames(folder_name):
image_list = [] image_list = []
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(folder_name)) EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(folder_name))
...@@ -240,9 +237,9 @@ def getExtractedFrames(folder_name): ...@@ -240,9 +237,9 @@ def getExtractedFrames(folder_name):
else: else:
return "No extracted frames were found" return "No extracted frames were found"
# get detections for a given frame name # get detections for a given frame name
def get_detections(video_name, frame_name): def get_detections(video_name, frame_name):
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name)) EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name))
FRAME_DIR = os.path.join(EXTRACTED_DIR, frame_name) FRAME_DIR = os.path.join(EXTRACTED_DIR, frame_name)
...@@ -257,7 +254,6 @@ def get_detections(video_name, frame_name): ...@@ -257,7 +254,6 @@ def get_detections(video_name, frame_name):
# get detections for a given class name # get detections for a given class name
def get_detections_for_label(video_name, label_index): def get_detections_for_label(video_name, label_index):
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name)) EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name))
CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5") CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5")
...@@ -328,7 +324,6 @@ def get_detections_for_label(video_name, label_index): ...@@ -328,7 +324,6 @@ def get_detections_for_label(video_name, label_index):
# to get the student evaluations # to get the student evaluations
def get_student_activity_evaluation(video_name): def get_student_activity_evaluation(video_name):
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name)) EXTRACTED_DIR = os.path.join(BASE_DIR, "assets\\FirstApp\\activity\\{}".format(video_name))
# CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5") # CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5")
...@@ -401,7 +396,6 @@ def get_frame_activity_recognition(video_name): ...@@ -401,7 +396,6 @@ def get_frame_activity_recognition(video_name):
# CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5") # CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_02.h5")
CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_04.h5") CLASSIFIER_DIR = os.path.join(BASE_DIR, "FirstApp\\classifiers\\student_activity_version_04.h5")
np.set_printoptions(suppress=True) np.set_printoptions(suppress=True)
# load the model # load the model
...@@ -473,7 +467,6 @@ def get_frame_activity_recognition(video_name): ...@@ -473,7 +467,6 @@ def get_frame_activity_recognition(video_name):
# increment the detection count # increment the detection count
detection_count += 1 detection_count += 1
# calculating the percentages for the frame # calculating the percentages for the frame
phone_checking_perct = float(phone_checking_count / detection_count) * 100 if detection_count > 0 else 0 phone_checking_perct = float(phone_checking_count / detection_count) * 100 if detection_count > 0 else 0
listening_perct = float(listening_count / detection_count) * 100 if detection_count > 0 else 0 listening_perct = float(listening_count / detection_count) * 100 if detection_count > 0 else 0
...@@ -575,7 +568,6 @@ def get_individual_student_evaluation(video_name, student_name): ...@@ -575,7 +568,6 @@ def get_individual_student_evaluation(video_name, student_name):
# this method will retrieve student activity summary for given time period # this method will retrieve student activity summary for given time period
def get_student_activity_summary_for_period(activities): def get_student_activity_summary_for_period(activities):
# declare variables to add percentage values # declare variables to add percentage values
phone_checking_perct_combined = 0.0 phone_checking_perct_combined = 0.0
listening_perct_combined = 0.0 listening_perct_combined = 0.0
...@@ -590,7 +582,6 @@ def get_student_activity_summary_for_period(activities): ...@@ -590,7 +582,6 @@ def get_student_activity_summary_for_period(activities):
# iterate through the activities # iterate through the activities
for activity in activities: for activity in activities:
individual_activity = {} individual_activity = {}
individual_activity["phone_perct"] = float(activity['phone_perct']) individual_activity["phone_perct"] = float(activity['phone_perct'])
individual_activity["listening_perct"] = float(activity['listening_perct']) individual_activity["listening_perct"] = float(activity['listening_perct'])
...@@ -603,7 +594,6 @@ def get_student_activity_summary_for_period(activities): ...@@ -603,7 +594,6 @@ def get_student_activity_summary_for_period(activities):
# append to the list # append to the list
individual_lec_activities.append(individual_activity) individual_lec_activities.append(individual_activity)
# calculate the average percentages # calculate the average percentages
phone_checking_average_perct = round((phone_checking_perct_combined / no_of_activities), 1) phone_checking_average_perct = round((phone_checking_perct_combined / no_of_activities), 1)
listening_average_perct = round((listening_perct_combined / no_of_activities), 1) listening_average_perct = round((listening_perct_combined / no_of_activities), 1)
...@@ -656,7 +646,6 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -656,7 +646,6 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict):
# assign the difference # assign the difference
frame_group_diff[key] = diff if diff > 0 else 1 frame_group_diff[key] = diff if diff > 0 else 1
# looping through the frames # looping through the frames
for frame in os.listdir(EXTRACTED_DIR): for frame in os.listdir(EXTRACTED_DIR):
# getting the frame folder # getting the frame folder
...@@ -682,7 +671,6 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -682,7 +671,6 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict):
image_array = np.asarray(image) image_array = np.asarray(image)
normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1 normalized_image_array = (image_array.astype(np.float32) / 127.0) - 1
# Load the image into the array # Load the image into the array
data[0] = normalized_image_array data[0] = normalized_image_array
...@@ -700,11 +688,9 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -700,11 +688,9 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict):
elif label == class_labels[2]: elif label == class_labels[2]:
note_count += 1 note_count += 1
# increment the detection count # increment the detection count
detection_count += 1 detection_count += 1
# finding the time landmark that the current frame is in # finding the time landmark that the current frame is in
for i in frame_landmarks: for i in frame_landmarks:
index = frame_landmarks.index(i) index = frame_landmarks.index(i)
...@@ -718,14 +704,11 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -718,14 +704,11 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict):
if (frame_count >= i) & (frame_count < next_value): if (frame_count >= i) & (frame_count < next_value):
frame_name = "{}-{}".format(i, next_value) frame_name = "{}-{}".format(i, next_value)
frame_group_dict[frame_name]['phone_count'] += phone_count frame_group_dict[frame_name]['phone_count'] += phone_count
frame_group_dict[frame_name]['listen_count'] += listen_count frame_group_dict[frame_name]['listen_count'] += listen_count
frame_group_dict[frame_name]['note_count'] += note_count frame_group_dict[frame_name]['note_count'] += note_count
frame_group_dict[frame_name]['detection_count'] += detection_count frame_group_dict[frame_name]['detection_count'] += detection_count
# increment the frame count # increment the frame count
frame_count += 1 frame_count += 1
...@@ -764,6 +747,83 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict): ...@@ -764,6 +747,83 @@ def activity_frame_groupings(video_name, frame_landmarks, frame_group_dict):
# print('frame group dict: ', frame_group_dict) # print('frame group dict: ', frame_group_dict)
activity_labels = ['phone_perct', 'listen_perct', 'note_perct'] activity_labels = ['phone_perct', 'listen_perct', 'note_perct']
# return the dictionary # return the dictionary
return frame_group_dict, activity_labels return frame_group_dict, activity_labels
\ No newline at end of file
# this section will handle saving activity entities to the database
def save_frame_recognition(video_name):
# retrieve the lecture activity id
lec_activity = LectureActivity.objects.filter(lecture_video_id__video_name=video_name)
lec_activity_ser = LectureActivitySerializer(lec_activity, many=True)
lec_activity_data = lec_activity_ser.data[0]
lec_activity_id = lec_activity_data['id']
# create a new lecture activity frame detections id
last_lec_activity_frame_recognitions = LectureActivityFrameRecognitions.objects.order_by(
'lecture_activity_frame_recognition_id').last()
new_lecture_activity_frame_recognitions_id = "LAFR00001" if (last_lec_activity_frame_recognitions is None) else \
ig.generate_new_id(last_lec_activity_frame_recognitions.lecture_activity_frame_recognition_id)
# calculate the frame detections
frame_detections = get_frame_activity_recognition(video_name)
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_activity_frame_recognition_details = LectureActivityFrameRecognitionDetails()
lec_activity_frame_recognition_details.frame_name = detection['frame_name']
lec_activity_frame_recognition_details.phone_perct = detection['phone_perct']
lec_activity_frame_recognition_details.listen_perct = detection['listening_perct']
lec_activity_frame_recognition_details.note_perct = detection['note_perct']
frame_recognition_details.append(lec_activity_frame_recognition_details)
lec_activity_frame_recognitions = LectureActivityFrameRecognitions()
lec_activity_frame_recognitions.lecture_activity_frame_recognition_id = new_lecture_activity_frame_recognitions_id
lec_activity_frame_recognitions.lecture_activity_id_id = lec_activity_id
lec_activity_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_activity_frame_recognitions.save()
# now return the frame detections
return frame_detections
# this method will save the activity frame groupings to the database
def save_frame_groupings(video_name, frame_landmarks, frame_group_dict):
frame_group_percentages, activity_labels = activity_frame_groupings(video_name, frame_landmarks,
frame_group_dict)
# save the frame group details into db
last_lec_activity_frame_grouping = LectureActivityFrameGroupings.objects.order_by(
'lecture_activity_frame_groupings_id').last()
new_lecture_activity_frame_grouping_id = "LAFG00001" if (last_lec_activity_frame_grouping is None) else \
ig.generate_new_id(last_lec_activity_frame_grouping.lecture_activity_frame_groupings_id)
# retrieve the lecture activity id
lec_activity = LectureActivity.objects.filter(lecture_video_id__video_name=video_name)
lec_activity_ser = LectureActivitySerializer(lec_activity, many=True)
lec_activity_id = lec_activity_ser.data[0]['id']
# create the frame group details
frame_group_details = []
for key in frame_group_percentages.keys():
# create an object of type 'LectureActivityFrameGroupDetails'
lec_activity_frame_group_details = LectureActivityFrameGroupDetails()
lec_activity_frame_group_details.frame_group = key
lec_activity_frame_group_details.frame_group_percentages = frame_group_percentages[key]
frame_group_details.append(lec_activity_frame_group_details)
new_lec_activity_frame_groupings = LectureActivityFrameGroupings()
new_lec_activity_frame_groupings.lecture_activity_frame_groupings_id = new_lecture_activity_frame_grouping_id
new_lec_activity_frame_groupings.lecture_activity_id_id = lec_activity_id
new_lec_activity_frame_groupings.frame_group_details = frame_group_details
# save
new_lec_activity_frame_groupings.save()
FirstApp/logic/head_gaze_estimation.py
View file @ 765ece4e
...@@ -16,6 +16,10 @@ import os ...@@ -16,6 +16,10 @@ import os
import shutil import shutil
import math import math
from ..MongoModels import *
from ..serializers import *
from . import id_generator as ig
def get_2d_points(img, rotation_vector, translation_vector, camera_matrix, val): def get_2d_points(img, rotation_vector, translation_vector, camera_matrix, val):
"""Return the 3D points present as 2D for making annotation box""" """Return the 3D points present as 2D for making annotation box"""
...@@ -846,3 +850,84 @@ def gaze_estimation_frame_groupings(video_name, frame_landmarks, frame_group_dic ...@@ -846,3 +850,84 @@ def gaze_estimation_frame_groupings(video_name, frame_landmarks, frame_group_dic
# return the dictionary # return the dictionary
return frame_group_dict, labels return frame_group_dict, labels
# this section will handle some database operations
def save_frame_detections(video_name):
# retrieve the lecture emotion id
lec_gaze = LectureGazeEstimation.objects.filter(lecture_video_id__video_name=video_name)
lec_gaze_ser = LectureGazeEstimationSerializer(lec_gaze, many=True)
lec_gaze_data = lec_gaze_ser.data[0]
lec_gaze_id = lec_gaze_data['id']
# create a new lecture activity frame detections id
last_lec_gaze_frame_recognitions = LectureGazeFrameRecognitions.objects.order_by(
'lecture_gaze_frame_recognition_id').last()
new_lecture_gaze_frame_recognitions_id = "LGFR00001" if (
last_lec_gaze_frame_recognitions is None) else \
ig.generate_new_id(last_lec_gaze_frame_recognitions.lecture_gaze_frame_recognition_id)
# calculate the frame detections
frame_detections, frame_rate = get_lecture_gaze_esrimation_for_frames(video_name)
# to be added to the field 'frame_recognition_details' in the Lecture Gaze Frame Recordings
frame_recognition_details = []
# save the new lecture activity frame recognitions
for detection in frame_detections:
lec_gaze_frame_recognition_details = LectureGazeFrameRecognitionDetails()
lec_gaze_frame_recognition_details.frame_name = detection['frame_name']
lec_gaze_frame_recognition_details.upright_perct = detection['upright_perct']
lec_gaze_frame_recognition_details.upleft_perct = detection['upleft_perct']
lec_gaze_frame_recognition_details.downright_perct = detection['downright_perct']
lec_gaze_frame_recognition_details.downleft_perct = detection['downleft_perct']
lec_gaze_frame_recognition_details.front_perct = detection['front_perct']
frame_recognition_details.append(lec_gaze_frame_recognition_details)
lec_gaze_frame_recognitions = LectureGazeFrameRecognitions()
lec_gaze_frame_recognitions.lecture_gaze_frame_recognition_id = new_lecture_gaze_frame_recognitions_id
lec_gaze_frame_recognitions.lecture_gaze_id_id = lec_gaze_id
lec_gaze_frame_recognitions.frame_recognition_details = frame_recognition_details
lec_gaze_frame_recognitions.save()
# now return the frame recognitions
return frame_detections
# this method will save gaze frame groupings to the database
def save_frame_groupings(video_name, frame_landmarks, frame_group_dict):
frame_group_percentages, gaze_labels = gaze_estimation_frame_groupings(video_name, frame_landmarks,
frame_group_dict)
# save the frame group details into db
last_lec_gaze_frame_grouping = LectureGazeFrameGroupings.objects.order_by('lecture_gaze_frame_groupings_id').last()
new_lecture_gaze_frame_grouping_id = "LGFG00001" if (last_lec_gaze_frame_grouping is None) else \
ig.generate_new_id(last_lec_gaze_frame_grouping.lecture_gaze_frame_groupings_id)
# retrieve the lecture activity id
lec_gaze = LectureGazeEstimation.objects.filter(lecture_video_id__video_name=video_name)
lec_gaze_ser = LectureGazeEstimationSerializer(lec_gaze, many=True)
lec_gaze_id = lec_gaze_ser.data[0]['id']
# create the frame group details
frame_group_details = []
for key in frame_group_percentages.keys():
# create an object of type 'LectureActivityFrameGroupDetails'
lec_gaze_frame_group_details = LectureGazeFrameGroupDetails()
lec_gaze_frame_group_details.frame_group = key
lec_gaze_frame_group_details.frame_group_percentages = frame_group_percentages[key]
frame_group_details.append(lec_gaze_frame_group_details)
new_lec_gaze_frame_groupings = LectureGazeFrameGroupings()
new_lec_gaze_frame_groupings.lecture_gaze_frame_groupings_id = new_lecture_gaze_frame_grouping_id
new_lec_gaze_frame_groupings.lecture_gaze_id_id = lec_gaze_id
new_lec_gaze_frame_groupings.frame_group_details = frame_group_details
# save
new_lec_gaze_frame_groupings.save()
FirstApp/logic/video_extraction.py
View file @ 765ece4e
...@@ -3,6 +3,11 @@ import cv2 ...@@ -3,6 +3,11 @@ import cv2
import shutil import shutil
import datetime import datetime
from FirstApp.MongoModels import *
from FirstApp.serializers import *
from . import id_generator as ig
def VideoExtractor(request): def VideoExtractor(request):
BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
...@@ -192,4 +197,100 @@ def getFrameLandmarks(video_name, category): ...@@ -192,4 +197,100 @@ def getFrameLandmarks(video_name, category):
frame_group_dict[name] = {'upright_count': 0, 'upleft_count': 0, 'downright_count': 0, 'downleft_count': 0, frame_group_dict[name] = {'upright_count': 0, 'upleft_count': 0, 'downright_count': 0, 'downleft_count': 0,
'front_count': 0, 'detection_count': 0} 'front_count': 0, 'detection_count': 0}
return frame_landmarks, frame_group_dict return frame_landmarks, frame_group_dict
\ No newline at end of file
# this section will handle some database operations
def save_time_landmarks(video_name):
last_lec_video_time_landmarks = LectureVideoTimeLandmarks.objects.order_by('lecture_video_time_landmarks_id').last()
new_lecture_video_time_landmarks_id = "LVTL00001" if (last_lec_video_time_landmarks is None) else \
ig.generate_new_id(last_lec_video_time_landmarks.lecture_video_time_landmarks_id)
# retrieve lecture video details
lec_video = LectureVideo.objects.filter(video_name=video_name)
lec_video_ser = LectureVideoSerializer(lec_video, many=True)
lec_video_id = lec_video_ser.data[0]['id']
# save the landmark details in the db
time_landmarks = getTimeLandmarks(video_name)
db_time_landmarks = []
# loop through the time landmarks
for landmark in time_landmarks:
landmark_obj = Landmarks()
landmark_obj.landmark = landmark
db_time_landmarks.append(landmark_obj)
new_lec_video_time_landmarks = LectureVideoTimeLandmarks()
new_lec_video_time_landmarks.lecture_video_time_landmarks_id = new_lecture_video_time_landmarks_id
new_lec_video_time_landmarks.lecture_video_id_id = lec_video_id
new_lec_video_time_landmarks.time_landmarks = db_time_landmarks
new_lec_video_time_landmarks.save()
# this method will save frame landmarks to the database
def save_frame_landmarks(video_name):
# retrieve the previous lecture video frame landmarks details
last_lec_video_frame_landmarks = LectureVideoFrameLandmarks.objects.order_by(
'lecture_video_frame_landmarks_id').last()
new_lecture_video_frame_landmarks_id = "LVFL00001" if (last_lec_video_frame_landmarks is None) else \
ig.generate_new_id(last_lec_video_frame_landmarks.lecture_video_frame_landmarks_id)
frame_landmarks, frame_group_dict = getFrameLandmarks(video_name, "Activity")
# retrieve lecture video details
lec_video = LectureVideo.objects.filter(video_name=video_name)
lec_video_ser = LectureVideoSerializer(lec_video, many=True)
lec_video_id = lec_video_ser.data[0]['id']
# save the frame landmarks details into db
db_frame_landmarks = []
for landmark in frame_landmarks:
landmark_obj = Landmarks()
landmark_obj.landmark = landmark
db_frame_landmarks.append(landmark_obj)
new_lec_video_frame_landmarks = LectureVideoFrameLandmarks()
new_lec_video_frame_landmarks.lecture_video_frame_landmarks_id = new_lecture_video_frame_landmarks_id
new_lec_video_frame_landmarks.lecture_video_id_id = lec_video_id
new_lec_video_frame_landmarks.frame_landmarks = db_frame_landmarks
new_lec_video_frame_landmarks.save()
# now return the frame landmarks and the frame group dictionary
return frame_landmarks, frame_group_dict
# this method will retrieve the frame landmarks from the database
def get_frame_landmarks(video_name):
frame_landmarks = []
# retrieve frame landmarks from db
lec_video_frame_landmarks = LectureVideoFrameLandmarks.objects.filter(lecture_video_id__video_name=video_name)
lec_video_frame_landmarks_ser = LectureVideoFrameLandmarksSerializer(lec_video_frame_landmarks, many=True)
lec_video_frame_landmarks_data = lec_video_frame_landmarks_ser.data[0]
retrieved_frame_landmarks = lec_video_frame_landmarks_data["frame_landmarks"]
# creating a new list to display in the frontend
for landmark in retrieved_frame_landmarks:
frame_landmarks.append(landmark['landmark'])
# now return the frame landmarks
return frame_landmarks
\ No newline at end of file
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment