def __init__(self, projectID, workers=None):
self.projectID = projectID
self.workers = workers
self.fileManager = FM()
self.projFileManager = self.fileManager.retProjFileManager(projectID)
self.mlFileManager = self.fileManager.retMLFileManager()
def __init__(self, projectID, workers=None):
self.projectID = projectID
if not self._checkProjectID():
raise Exception(projectID + ' is not valid.')
self.workers = workers
self.fileManager = FM(projectID=projectID)
def __init__(self):
# 1: Define valid commands and ignore warnings
self.commands = [
'New', 'Restart', 'Stop', 'Rewrite', 'UploadData', 'LocalDelete',
'Snapshots'
]
np.seterr(invalid='ignore')
# 2: Determine which Kinect is attached (This script can handle v1 or v2 Kinects)
self._identifyDevice() #Stored in self.device
# 3: Create file manager
self.fileManager = FM()
# 4: Download credential files
self.fileManager.downloadData(self.localCredentialSpreadsheet)
self.credentialSpreadsheet = self.fileManager.localCredentialSpreadsheet # Rename to make code readable
# 5: Identify credential files (Credential files for uploading updates to Google Drive are found here)
self.credentialSpreadsheet = self.masterDirectory + 'CredentialFiles/SAcredentials.json'
# 6: Connect to Google Spreadsheets
self._authenticateGoogleSpreadSheets() #Creates self.controllerGS
self._modifyPiGS(error='')
# 7: Start PiCamera
try:
from picamera import PiCamera
self.camera = PiCamera()
self.camera.resolution = (1296, 972)
self.camera.framerate = 30
self.piCamera = 'True'
except Exception:
self.piCamera = 'False'
# 8: Keep track of processes spawned to convert and upload videofiles
self.processes = []
# 9: Await instructions
self.monitorCommands()
def __init__(self):
# 1: Define valid commands and ignore warnings
self.commands = [
'New', 'Restart', 'Stop', 'Rewrite', 'UploadData', 'LocalDelete',
'Snapshots'
]
np.seterr(invalid='ignore')
# 2: Determine which Kinect is attached (This script can handle v1 or v2 Kinects)
self._identifyDevice() #Stored in self.device
self.system = platform.node()
# 3: Create file manager
self.fileManager = FM()
# 4: Download credential files
self.fileManager.downloadData(
self.fileManager.localCredentialSpreadsheet)
self.fileManager.downloadData(self.fileManager.localCredentialDrive)
self.credentialSpreadsheet = self.fileManager.localCredentialSpreadsheet # Rename to make code readable
# 5: Connect to Google Spreadsheets
self._authenticateGoogleSpreadSheets() #Creates self.controllerGS
self._modifyPiGS(error='')
# 6: Start PiCamera
self.camera = PiCamera()
self.camera.resolution = (1296, 972)
self.camera.framerate = 30
self.piCamera = 'True'
# 7: Keep track of processes spawned to convert and upload videofiles
self.processes = []
# 8: Set size of frame
self.r = (0, 0, 640, 480)
# 9: Await instructions
self.monitorCommands()
from Modules.FileManagers.FileManager import FileManager as FM
import subprocess, pdb
pdb.set_trace()
anFM_obj = FM().retAnFileManager()
labeledClipsDir = anFM_obj.prepareVideoAnnotation('10classLabels')
#labeledClipsDir = '/Users/pmcgrath7/Temp/CichlidAnalyzer/__AnnotatedData/LabeledVideos/10classLabels/LabeledClips/'
#pdb.set_trace()
#subprocess.run(['python3', 'Modules/MachineLearning/3D_resnet.py', '--data', labeledClipsDir])
'--Practice',
action='store_true',
help='Use if you dont want to save annotations')
parser.add_argument('-i',
'--Initials',
type=str,
help='Initials to save annotations')
args = parser.parse_args()
if args.Initials is None:
initials = socket.gethostname()
else:
initials = args.Initials
fm_obj = FM(projectID=args.ProjectID)
fm_obj.createDirectory(fm_obj.localAnalysisDir)
fm_obj.downloadData(fm_obj.localManualLabelClipsDir, tarred=True)
fm_obj.downloadData(fm_obj.localLabeledClipsFile)
temp_csv = fm_obj.localAnalysisDir + 'NewAnnotations.csv'
# Read in annotations and create csv file for all annotations with the same user and projectID
dt = pd.read_csv(fm_obj.localLabeledClipsFile, index_col='LID')
new_dt = pd.DataFrame(columns=dt.columns)
clips = [
x for x in os.listdir(fm_obj.localManualLabelClipsDir)
if 'ManualLabel.mp4' in x
]
categories = ['c', 'f', 'p', 't', 'b', 'm', 's', 'x', 'o', 'd', 'q', 'k']
parser = argparse.ArgumentParser(
description='This command runs HMM analysis on a single row of data.')
parser.add_argument('ProjectID', type=str, help='ProjectID to analyze')
parser.add_argument('-n',
'--Number',
type=int,
help='Limit annotation to x number of frames.')
parser.add_argument('-p',
'--Practice',
action='store_true',
help='Use if you dont want to save annotations')
args = parser.parse_args()
fileManager = FM()
projFileManager = fileManager.retProjFileManager(args.ProjectID)
projFileManager.downloadData('ObjectLabeler')
anFileManager = fileManager.retAnFileManager()
obj = ObjectLabeler(projFileManager.localManualLabelFramesDir,
projFileManager.localLabeledFramesFile, args.Number,
args.ProjectID)
if not args.Practice:
# Backup annotations. Redownload to avoid race conditions
if not os.path.exists(projFileManager.localLabeledFramesFile):
print(projFileManager.localLabeledFramesFile +
'does not exist. Did you annotate any new frames? Quitting...')
else:
from Modules.FileManagers.FileManager import FileManager as FM
import subprocess
import pdb
fm_obj = FM()
pdb.set_trace()
a = fm_obj.localOrganizedLabeledClipsDir
b = fm_obj.local3DVideosDir
#fm_obj.downloadAnnotationData('LabeledVideos')
#subprocess.run(['python3', 'Modules/MachineLearning/3D_resnet.py', '--data', fm_obj.localOrganizedLabeledClipsDir, '--results', fm_obj.local3DVideosDir])
def __init__(self):
__version__ = '1.0.0'
self.fileManager = FM()
self.projectTypes = [
'Prep', 'Depth', 'Cluster', 'MLCluster', 'MLObject', 'Summary'
]
