DollarPedestrianDetectionCode - Revision history

Mtaiana: /* Getting ready */

2014-05-08T17:40:53Z

Getting ready

← Older revision		Revision as of 17:40, 8 May 2014
Line 17:		Line 17:
	# the res subdirectory, '''$code/data-INRIA/res''', containing more subdirectories with the results of the detections obtained running various algorithms		# the res subdirectory, '''$code/data-INRIA/res''', containing more subdirectories with the results of the detections obtained running various algorithms

	~~Annotations~~ are in the format: (x0, y0, deltaX, deltaY, confidence). The confidence value is used to plot the ROC curves.		The detections are stored in text files. There can be either one single file for one whole directory or one file for each image. In the first case the format for the file is: (image#, x0, y0, deltaX, deltaY, confidence). In the second case, the first field is missing. The confidence value is used to plot the ROC curves.
	~~There is one annotation file for each image~~.
	I don't know if the top-left corner is (0,0) or (1,1).		I don't know if the top-left corner is (0,0) or (1,1).

Mtaiana at 17:38, 8 May 2014

2014-05-08T17:38:15Z

← Older revision		Revision as of 17:38, 8 May 2014
Line 104:		Line 104:
	(Be careful to select only one algorithm to evaluate, otherwise you will create at least tens of albums!)		(Be careful to select only one algorithm to evaluate, otherwise you will create at least tens of albums!)

	plotBb( res, rDir, 30, 'fp' ); %The third parameter tells the system how many pages of ~~fals~~ positives to save at most		plotBb( res, rDir, 30, 'fp' ); %The third parameter tells the system how many pages of false positives to save at most
	plotBb( res, rDir, 0, 'tp' ); %If you want to save true positives		plotBb( res, rDir, 0, 'tp' ); %If you want to save true positives
	plotBb( res, rDir, 30, 'fn' ); %The third parameter tells the system how many pages of ~~fals~~ negatives to save at most		plotBb( res, rDir, 30, 'fn' ); %The third parameter tells the system how many pages of false negatives to save at most

	==Various==		==Various==

Mtaiana: /* Create new annotation (INRIA data set) */

2014-04-07T18:50:25Z

Create new annotation (INRIA data set)

← Older revision		Revision as of 18:50, 7 April 2014
Line 154:		Line 154:
	=== Create new annotation (INRIA data set) ===		=== Create new annotation (INRIA data set) ===
	It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.		It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.
	So we might do that using '''bblabeler.m''', which saves the annotations in a different format (one file per image) ~~, but should be useful anyway.~~		So we might do that using '''bblabeler.m''', which saves the annotations in a different format (one file per image).
	~~I think it's possible to then read all the labels via vbbload and write them as one vbb file~~.

	bbLabeler( [{'person', 'people'}], '~/PhD/Datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');		bbLabeler( [{'person', 'people'}], '~/PhD/Datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');

			The single annotation files are combined into one single vbb file by running:

			[B] = vbb( 'vbbFrFiles', 'NewAnnotations' )
			vbb('vbbSave', B, 'V000.vbb')

	=== Transform all the images of a data set into a new version of it ===		=== Transform all the images of a data set into a new version of it ===
	Use the "convert" function on the seq file, applying imgFun(I) to each frame I:		Use the "convert" function on the seq file, applying imgFun(I) to each frame I:
	seqIo( fName, 'convert', tName, imgFun, varargin )		seqIo( fName, 'convert', tName, imgFun, varargin )

Mtaiana: /* Annotating the INRIA data set again */

2014-04-07T18:16:59Z

Annotating the INRIA data set again

← Older revision		Revision as of 18:16, 7 April 2014
Line 152:		Line 152:
	imshow(image);		imshow(image);

	=== ~~Annotating the~~ INRIA data set ~~again~~ ===		=== Create new annotation (INRIA data set) ===
	It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.		It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.
	So we might do that using '''bblabeler.m''', which saves the annotations in a different format, but should be useful anyway.		So we might do that using '''bblabeler.m''', which saves the annotations in a different format (one file per image) , but should be useful anyway.
			I think it's possible to then read all the labels via vbbload and write them as one vbb file.

	bbLabeler( [{'person', 'people'}], '~/PhD/~~datasets~~/INRIAPerson/Train/pos/', 'inriaNewLabels');		bbLabeler( [{'person', 'people'}], '~/PhD/Datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');

	=== Transform all the images of a data set into a new version of it ===		=== Transform all the images of a data set into a new version of it ===
	Use the "convert" function on the seq file, applying imgFun(I) to each frame I:		Use the "convert" function on the seq file, applying imgFun(I) to each frame I:
	seqIo( fName, 'convert', tName, imgFun, varargin )		seqIo( fName, 'convert', tName, imgFun, varargin )

Mtaiana at 16:00, 6 February 2014

2014-02-06T16:00:07Z

← Older revision		Revision as of 16:00, 6 February 2014
Line 108:		Line 108:
	plotBb( res, rDir, 30, 'fn' ); %The third parameter tells the system how many pages of fals negatives to save at most		plotBb( res, rDir, 30, 'fn' ); %The third parameter tells the system how many pages of fals negatives to save at most

			==Various==
	=== Read one or more imags from a ".seq" file and write it to a regular file ===		=== Read one or more imags from a ".seq" file and write it to a regular file ===
	Is = seqIo( 'data-INRIA/videos/set00/V000.seq', 'toImgs', '.', 1, 0, 2, 'png' )		Is = seqIo( 'data-INRIA/videos/set00/V000.seq', 'toImgs', '.', 1, 0, 2, 'png' )
Line 156:		Line 157:

	bbLabeler( [{'person', 'people'}], '~/PhD/datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');		bbLabeler( [{'person', 'people'}], '~/PhD/datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');

			=== Transform all the images of a data set into a new version of it ===
			Use the "convert" function on the seq file, applying imgFun(I) to each frame I:
			seqIo( fName, 'convert', tName, imgFun, varargin )

Anambiar at 14:26, 11 July 2013

2013-07-11T14:26:10Z

@@ Line 111: / Line 111: @@
    Is = seqIo( 'data-INRIA/videos/set00/V000.seq', 'toImgs', '.',              1,                  0,               2,          'png' )
                 input seq file                      command  dest directory    frames to skip      first frame   last frame     format
 === Adding a new dataset WIP ===

Mtaiana: /* Adding a new dataset WIP */

2012-03-05T11:13:10Z

Adding a new dataset WIP

← Older revision		Revision as of 11:13, 5 March 2012
Line 125:		Line 125:
	vidIds={0}; %video ID's inside each of the sub datasets		vidIds={0}; %video ID's inside each of the sub datasets

	To create a seq file staring from a video, you can use the [http://users.isr.ist.utl.pt/~mtaiana/code/createSeqFile.m createSeqFile.m] script. For some reason, the seq files created this way are not playable directly with '''vbbPlayer.m''', but it can be labelled using '''vbbLabeler.m'''.		To create a seq file staring from a video, you can use the [http://users.isr.ist.utl.pt/~mtaiana/code/createSeqFile.m createSeqFile.m] script. For some reason, the seq files created this way are not playable directly with '''vbbPlayer.m''', but it can be labelled using '''vbbLabeler.m'''. You can download a short tutorial on how to use the labeller here: [http://users.isr.ist.utl.pt/~mtaiana/temp/LabellingDemo.mkv labelling demo].

	The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.		The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.

Mtaiana: /* Adding a new dataset WIP */

2012-02-09T13:11:04Z

Adding a new dataset WIP

← Older revision		Revision as of 13:11, 9 February 2012
Line 125:		Line 125:
	vidIds={0}; %video ID's inside each of the sub datasets		vidIds={0}; %video ID's inside each of the sub datasets

	To create a seq file staring from a video, you can use the ~~'''~~createSeqFile.m~~'''~~ script. For some reason, the seq files created this way are not playable directly with '''vbbPlayer.m''', but it can be labelled using '''vbbLabeler.m'''.		To create a seq file staring from a video, you can use the [http://users.isr.ist.utl.pt/~mtaiana/code/createSeqFile.m createSeqFile.m] script. For some reason, the seq files created this way are not playable directly with '''vbbPlayer.m''', but it can be labelled using '''vbbLabeler.m'''.

	The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.		The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.

Mtaiana at 18:46, 6 February 2012

2012-02-06T18:46:54Z

← Older revision		Revision as of 18:46, 6 February 2012
Line 128:		Line 128:

	The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.		The classes for the bounding boxes are '''person''' for a single person, '''person?''' for something the labeler is not sure whether it is a person or not, '''people''' for groups of people for which it is impossible to label each person correctly. There is a label for persons who are partly occluded, it's not so clear how it works.

			=== Accessing the frames of a seq file ===

			%Initialize the reader
			seqReader = seqIo( 'cam60.seq', 'reader');
			%Read one frame
			seqReader.seek(269);
			image = seqReader.getframe();
			imshow(image);

	=== Annotating the INRIA data set again ===		=== Annotating the INRIA data set again ===

Mtaiana: /* Annotating the INRIA data set again */

2012-01-24T19:31:41Z

Annotating the INRIA data set again

← Older revision		Revision as of 19:31, 24 January 2012
Line 132:		Line 132:
	It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.		It seems impossible to do it with '''vbbLabeler.m': it complains about something, like the fact that the images are not all of the same size.
	So we might do that using '''bblabeler.m''', which saves the annotations in a different format, but should be useful anyway.		So we might do that using '''bblabeler.m''', which saves the annotations in a different format, but should be useful anyway.

			bbLabeler( [{'person', 'people'}], '~/PhD/datasets/INRIAPerson/Train/pos/', 'inriaNewLabels');