[Zdenek Kalal, et al., PAMI Jan, 2010] Tracking-Learning-Detection

Contribution:

• A tracking framework (TLD)
• decomposes the long-term tracking task into tracking, learning, detection
• A learning method (P-N learning)
• estimates the errors by a pair of "experts"
• P-expert: estimates missed detections
• N-expert: estimates false alarms 
• Real-time implementation of TLD and P-N learning

Literature Review 2

From George:

Please find bellow some references:

 

---------------------------------------------------------------------------------------------------------------------------

- A paper on wearable eye tracking. You may be able to find more references, but this is a good starting point

 

http://thirtysixthspan.com/openEyes/li_etal06.pdf

 

and a quite extensive thesis with a lot of information on eye tracking (among other irrelevant material)

 

http://edoc.ub.uni-muenchen.de/11591/1/Drewes_Heiko.pdf

 

 

 

 

---------------------------------------------------------------------------------------------------------------------------

- The following are related to the last part of your project, when your 3D eye tracking and basic 3D object tracking are in place.

 

These are interesting from a technical point of view (should open the full papers from these links on IC computer):

 

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=06634139

 

http://stefan.winkler.net/Publications/qomex2013eye.pdf

 

With a useful presentationhttp://www.geo.uzh.ch/microsite/icacogvis/pdf/icc2013/Kurzhals.pdf

 

Another summaryhttp://docs.alexanno.net/files/ICA_ICC_nossum_opach_poster_abstract_final.pdf

 

And these are interesting from a clinical point of view, although a screen-based eye-tracker is in use here and also the context is slightly different to what you are trying to achieve:

 

Orientation strategies in natural orifice translumenal endoscopic surgery.

Sodergren MH, Orihuela-Espina F, Mountney P, Clark J, Teare J, Darzi A, Yang GZ.

Ann Surg. 2011 Aug;254(2):257-66. doi: 10.1097/SLA.0b013e31822513c6.

 

Evaluation of orientation strategies in laparoscopic cholecystectomy.

Sodergren MH, Orihuela-Espina F, Clark J, Teare J, Yang GZ, Darzi A.

Ann Surg. 2010 Dec;252(6):1027-36. doi: 10.1097/SLA.0b013e3181e49683.

 

 

Some more ideas on how for instance multi-person 3D eye-tracking in a surgical theatre could be used to facilitate collaboration among the clinical staff.

 

Chetwood AS, Kwok KW, Sun LW, et al., 2012, Collaborative eye tracking: a potential training tool in laparoscopic surgery., Surg Endosc, Vol:26, 0930-2794, Pages:2003-2009

• Author Web Link

Kwok KW, Sun LW, Mylonas GP, et al., 2012, Collaborative gaze channelling for improved cooperation during robotic assisted surgery., Ann Biomed Eng, Vol:40, 0090-6964, Pages:2156-2167

• Author Web Link

Literature Review 1

From George:

Please find below some papers related to the first task of your project as discussed during our meeting. This concerns the recovery of the eye-tracker’s scene-camera pose by using the 2D image and the Kinect provided image and depth map. Not all of the material may be directly relevant but nevertheless will give you a good starting point. Whenever in doubt, please ask.

http://epubs.surrey.ac.uk/713800/1/Kalal-PAMI-2011(1).pdf

http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6175907

http://cvlabwww.epfl.ch/~lepetit/papers/lepetit_ijcv08.pdf

I will give you on time maybe some papers related to how a monocular/wearable eye-tracker works, some papers for collaboration through eye-tracking and finally existing papers on workflow segmentation in the surgical environment.