From awd at cs.cmu.edu Mon Feb 2 11:20:15 2009 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Mon, 02 Feb 2009 11:20:15 -0500 Subject: [Research] Auton Lab meeting Tuesday February 3rd 12:00 *NOON* NSH 1507 Message-ID: <49871D3F.9070901@cs.cmu.edu> Dear Autonians, We will have a post-superbowl-post-groundhog-day meeting tomorrow. It will be led by Michael Knight. A hundred yard interception->touchdown run is nothing in comparison to what he has to show. Food will be provided. Please note the new starting time of our meetings: from now on, they will start at noon. See you tomorrow, Artur From awd at cs.cmu.edu Fri Feb 6 10:33:28 2009 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Fri, 06 Feb 2009 10:33:28 -0500 Subject: [Research] Auton Lab meeting: *THURSDAY* February 12th, 11:30 @ NSH1507 Message-ID: <498C5848.4050202@cs.cmu.edu> Speaker: Sajid Siddiqi, ABD Topic: TBA Food: Yes. Please note that this meeting will be taking place on Thursday, and that we won't have a lab meeting on Tuesday Feb 10th. Artur From awd at cs.cmu.edu Fri Feb 6 15:19:12 2009 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Fri, 06 Feb 2009 15:19:12 -0500 Subject: [Research] DATE/TIME/PLACE CHANGE Re: Auton Lab meeting: *THURSDAY* February 12th, 11:30 @ NSH1507 In-Reply-To: <498C5848.4050202@cs.cmu.edu> References: <498C5848.4050202@cs.cmu.edu> Message-ID: <498C9B40.1030407@cs.cmu.edu> Dear Autonians, We had to move this meeting to *Friday Feb 13, 3:30pm*. It will take place in *NSH 3305* and there will be no food at this meeting due to the awkward time of it. Below I am pasting the title and summary of Sajid's talk (at least the speaker stayed the same :) Please attend if you can. This will be a practice job talk that Sajid will be giving at JPL the following week. Thanks Artur Title: Learning Latent Variable Models of Dynamical Systems Abstract: Dynamical systems and time series data are ubiquitous in our world. Models for dynamical systems find widespread application for understanding these systems, predicting their future behavior, time series classification and recognition, planning and control, and much more. In this talk I will give an overview of a popular class of dynamical system models, namely latent variable models, with an emphasis on Hidden Markov Models (HMMs) and Linear Dynamical Systems (LDS). I will describe work we have done on novel learning algorithms for HMMs and LDSs that are more efficient, accurate and robust than their predecessors in many ways. We will also discuss recent work on a hybrid latent variable model that captures aspects of both HMMs and LDSs. We will examine results on human motion capture data for graphics, dynamic texture videos for computer vision, and mobile robot camera data for robotics applications. Artur Dubrawski wrote: > Speaker: Sajid Siddiqi, ABD > Topic: TBA > Food: Yes. > > Please note that this meeting will be taking place on Thursday, > and that we won't have a lab meeting on Tuesday Feb 10th. > > Artur > > _______________________________________________ > Research mailing list > Research at autonlab.org > https://www.autonlab.org/mailman/listinfo/research > > From psarkar at cs.cmu.edu Sat Feb 7 22:34:05 2009 From: psarkar at cs.cmu.edu (Purnamrita Sarkar) Date: Sat, 7 Feb 2009 22:34:05 -0500 (EST) Subject: [Research] Latex help! Message-ID: <62818.68.162.181.233.1234064045.squirrel@webmail.cs.cmu.edu> Hi All, I have been using JPG figures for a paper. Pdflatex does a fine job to get tex->pdf straight. But now, I see that the conference needs PS files to be submitted as well. The problem is if I try Latex->dvi->ps then it doesn't compile with JPG files. Complains that jpg files don't ahve a bounding box,, etc. Is there any way to get a ps file with Jpg figures from Latex. I tried some run-of-the-mill pdf-to-ps converters, but the results are Awful :(! Any ideas? Thanks!!! Purna From komarek.paul at gmail.com Sun Feb 8 02:41:26 2009 From: komarek.paul at gmail.com (Paul Komarek) Date: Sat, 7 Feb 2009 23:41:26 -0800 Subject: [Research] Latex help! In-Reply-To: <62818.68.162.181.233.1234064045.squirrel@webmail.cs.cmu.edu> References: <62818.68.162.181.233.1234064045.squirrel@webmail.cs.cmu.edu> Message-ID: <459f38470902072341u2c0f7852wcc15e9e2c36184b4@mail.gmail.com> Good graphics in PS need vector image formats, hence jpeg will never work well. If your jpeg is "screen resolution", then the conversion to PS will look good on the screen, but not on the printer. And vice-versa. How are you generating your jpeg originals? Paul On Sat, Feb 7, 2009 at 7:34 PM, Purnamrita Sarkar wrote: > Hi All, > > I have been using JPG figures for a paper. Pdflatex does a fine job to get > tex->pdf straight. But now, I see that the conference needs PS files to be > submitted as well. The problem is if I try Latex->dvi->ps then it doesn't > compile with JPG files. Complains that jpg files don't ahve a bounding > box,, etc. > > Is there any way to get a ps file with Jpg figures from Latex. I tried > some run-of-the-mill pdf-to-ps converters, but the results are Awful :(! > Any ideas? > > Thanks!!! > Purna > > _______________________________________________ > Research mailing list > Research at autonlab.org > https://www.autonlab.org/mailman/listinfo/research > From psarkar at cs.cmu.edu Sun Feb 8 15:19:51 2009 From: psarkar at cs.cmu.edu (Purnamrita Sarkar) Date: Sun, 08 Feb 2009 15:19:51 -0500 Subject: [Research] Latex help! In-Reply-To: <459f38470902072341u2c0f7852wcc15e9e2c36184b4@mail.gmail.com> References: <62818.68.162.181.233.1234064045.squirrel@webmail.cs.cmu.edu> <459f38470902072341u2c0f7852wcc15e9e2c36184b4@mail.gmail.com> Message-ID: <498F3E67.3090603@cs.cmu.edu> Hi All, Thanks for all the helpful ideas! Gimp does work reasonably well to convert to eps format. But now, when I print the PS file (even if all the figures look fine) the head gets cut off! This is only for the ps file, not the pdf. Any ideas? Thanks again! Purna From awd at cs.cmu.edu Fri Feb 13 08:59:45 2009 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Fri, 13 Feb 2009 08:59:45 -0500 Subject: [Research] Reminder: Practice Job Talk Today 3:30pm NSH 3305 Message-ID: <49957CD1.8080809@cs.cmu.edu> -------------- next part -------------- An embedded message was scrubbed... From: "Sajid Siddiqi" Subject: please post to research at autonlab.org Date: Fri, 13 Feb 2009 08:42:59 -0500 (EST) Size: 3021 URL: From sajid at cmu.edu Fri Feb 13 14:12:17 2009 From: sajid at cmu.edu (Sajid M. Siddiqi) Date: Fri, 13 Feb 2009 14:12:17 -0500 (EST) Subject: [Research] Reminder: Practice Job Talk Today 3:30pm NSH 3305 Message-ID: <3631.128.237.247.174.1234552337.squirrel@128.237.247.174> Hi all, Please come and share your feedback with me for my practice talk today, in NSH 3305 at 3:30pm. Your input will be greatly appreciated. Thanks! Title: Learning Latent Variable Models of Dynamical Systems Abstract: Dynamical systems and time series data are ubiquitous in our world. Models for dynamical systems find widespread application for understanding these systems, predicting their future behavior, time series classification and recognition, planning and control, and much more. In this talk I will give an overview of a popular class of dynamical system models, namely latent variable models, with an emphasis on Hidden Markov Models (HMMs) and Linear Dynamical Systems (LDS). I will describe work we have done on novel learning algorithms for HMMs and LDSs that are more efficient, accurate and robust than their predecessors in many ways. In particular, I will focus on the problem of learning models of appropriate dimensionality, on learning stable dynamics, and on learning the underlying manifold of a dynamical system. I also briefly describe recent work on a hybrid latent variable model that captures aspects of both HMMs and LDSs. I will present results on human motion capture data for graphics, dynamic texture videos for computer vision, and mobile robot camera data for robotics applications. Sajid From awd at cs.cmu.edu Tue Feb 17 09:27:40 2009 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Tue, 17 Feb 2009 09:27:40 -0500 Subject: [Research] no Auton Lab meeting today Message-ID: <499AC95C.1010604@cs.cmu.edu> From tzukuoh at cs.cmu.edu Wed Feb 25 15:53:06 2009 From: tzukuoh at cs.cmu.edu (Tzu-Kuo Huang) Date: Wed, 25 Feb 2009 15:53:06 -0500 Subject: [Research] FW: Yi Zhang wins IBM Fellowship! Message-ID: <002101c9978b$0f923c20$2eb6b460$@cmu.edu> Congratulations Yi!! -----Original Message----- From: ml-phd-students-bounces at LOGANBERRY.srv.cs.cmu.edu [mailto:ml-phd-students-bounces at LOGANBERRY.srv.cs.cmu.edu] On Behalf Of Tom Mitchell Sent: Wednesday, February 25, 2009 3:38 PM To: ml-faculty at cs.cmu.edu; ml-students at cs.cmu.edu Subject: Yi Zhang wins IBM Fellowship! I'm very pleased to announce that our own Yi Zhang has just been awarded an IBM Fellowship. As many of you know, this is a highly competitive fellowship, and receiving it is a true honor. Congratulations Yi! Tom -- Tom M. Mitchell Fredkin Professor and Department Head Machine Learning Department Carnegie Mellon University www.cs.cmu.edu/~tom From sabhnani at cs.cmu.edu Wed Feb 25 16:26:25 2009 From: sabhnani at cs.cmu.edu (Robin Sabhnani) Date: Wed, 25 Feb 2009 16:26:25 -0500 (EST) Subject: [Research] FW: Yi Zhang wins IBM Fellowship! In-Reply-To: <002101c9978b$0f923c20$2eb6b460$@cmu.edu> References: <002101c9978b$0f923c20$2eb6b460$@cmu.edu> Message-ID: <41863.128.2.181.158.1235597185.squirrel@webmail.cs.cmu.edu> Heartly congrats to Yi !! On Wed, February 25, 2009 3:53 pm, Tzu-Kuo Huang wrote: > Congratulations Yi!! > > > -----Original Message----- > From: ml-phd-students-bounces at LOGANBERRY.srv.cs.cmu.edu > [mailto:ml-phd-students-bounces at LOGANBERRY.srv.cs.cmu.edu] On Behalf Of > Tom > Mitchell > Sent: Wednesday, February 25, 2009 3:38 PM > To: ml-faculty at cs.cmu.edu; ml-students at cs.cmu.edu > Subject: Yi Zhang wins IBM Fellowship! > > > I'm very pleased to announce that our own Yi Zhang has just been > awarded an IBM Fellowship. As many of you know, this is a highly > competitive fellowship, and receiving it is a true honor. Congratulations > Yi! > > > Tom > > > -- > Tom M. Mitchell > Fredkin Professor and Department Head > Machine Learning Department > Carnegie Mellon University > www.cs.cmu.edu/~tom > > > _______________________________________________ > Research mailing list > Research at autonlab.org > https://www.autonlab.org/mailman/listinfo/research > > > -- Robin Sabhnani Machine Learning PhD student From schneide at cs.cmu.edu Fri Feb 27 20:09:19 2009 From: schneide at cs.cmu.edu (Jeff Schneider) Date: Fri, 27 Feb 2009 20:09:19 -0500 Subject: [Research] Auton Lab meeting Tuesday March 3rd 12:00 *NOON* In-Reply-To: <49871D3F.9070901@cs.cmu.edu> References: <49871D3F.9070901@cs.cmu.edu> Message-ID: <49A88EBF.7060204@cs.cmu.edu> There will be an Auton Lab meeting this Tuesday at noon. More official details to follow. TK will lead a discussion on his latest work (and ICML submission) on learning dynamical models from data that is not sequenced or time-stamped (i.e. is not in the form of a trajectory of observations). Along the way we will address questions such as: Are you crazy, how can this be possible? Why on earth would you want to do this? (hint: think of things not on earth)