Lourdes Agapito

I hold the position of Professor of 3D Vision in the Department of Computer Science at University College London (UCL).

My research in Computer Vision has consistently focused on the inference of 3D information from the video acquired from a single moving camera. While my early research focused on static scenes, my attention soon turned to the much more challenging problem of estimating the 3D shape of non-rigid objects (Non-Rigid Structure from Motion, NR-SFM) or complex dynamic scenes where an unknown number of objects might be moving, possibly deforming, independently. My research group investigates all theoretical and practical aspects of NRSFM: deformable tracking; dense optical flow estimation and non-rigid video registration; 3D reconstruction of deformable and articulated structure and dense 3D modelling of non-rigid dynamic scenes.

I have held an ERC Starting Grant funded by the European Research Council from 2008-2014. I am a member of the Vision and Imaging Science group and the Centre for Inverse Problems.

RECENT NEWS

• January 2015: I am Program Chair for CVPR'16.
• January 2015: I am hiring 1 Postdoc on my new EU grant.
• January 2015: I have 3 PhD studentships available on funded projects (only for UK/EU students) starting in May 2015. I am looking for outstanding candidates with an MSc in Computer Vision (distinction).
• December 2014: I have just returned from attending 3DV'14 in Tokyo. It was a great conference and a very rewarding experience to have been Program Chair with Martial Hebert and Hiroshi Kawasaki.
• September 2014: Our EU Horizon 2020 proposal Second Hands has been accepted for funding. This is a 5 year grant with Ocado Technology, EPFL, KIT and UniRoma to design and build a robot that will assist a human in maintenance tasks. My group will be doing the 3D visual perception for the robot.
• June 2014: We just returned from CVPR'14 where our oral paper Reconstructing Pascal VOC received a lot of attention. We have been invited to submit an extended version to the PAMI special issue on the best papers from CVPR'14.
• June 2014: Our paper "Video-Popup: Monocular 3D Reconstruction of Dynamic Scenes" has been accepted as an oral presentation to ECCV'14!
• March 2014: We have had two papers accepted at CVPR'14 (one oral).
  
• August 2013: I am Program Chair of the International Conference on 3D Vision (3DV'14) which will take place in Tokyo in December 2014.
  
• July 2013: I have moved to University College London (UCL) where I have taken up the position of Reader in Computer Vision at the Computer Science Department.
  
• June 2013: I will serve as Area Chair for ECCV 2014.
  
• June 2013: We just returned from CVPR'13. Ravi's oral presentation of our work on Dense Reconstruction of Non-Rigid Surfaces from Monocular Video received great attention. Great to see that Google Research referred to it in their post on Google+ of the highlights of the conference!
  
• May 2013: I will serve as Area Chair for CVPR 2014.
  
• April 2013: I will serve as Area Chair for ACCV 2014.
  
• March 2013: I am Workshop Chair for ECCV 2014.
  
• February 2013: 3 papers accepted to CVPR 2013.
  - A Variational Formulation for Dense Non Rigid Structure from Motion (oral 3% acceptance rate)
  Ravi Garg, Anastasios Roussos, Lourdes Agapito. See our video!
  - Learning Manifolds as Connected Linear Subspaces
  Nikolaos Pitelis, Chris Russell, Lourdes Agapito.
  - Looking Beyond the Image: Unsupervised Learning for Object Saliency and Detection
  Parthipan Siva, Chris Russell, Tao Xiang, Lourdes Agapito.
  
  

• December 2012: Our paper "A Variational Approach to Video Registration with Subspace Constraints" has been accepted to IJCV (see project webpage for videos and new dataset).

• September 2012: My PhD student Marco Paladini has been awarded the 2011 BMVA Sullivan Prize for his PhD Thesis with title "Deformable and Articulated 3D Reconstruction from Monocular Video Sequences". The Sullivan prize is awarded each year by the British Machine Vision Association to the best doctoral thesis submitted to a UK University, in the field of Computer Vision.

• August 2012: New papers accepted to ECCV'12 and ISMAR'12.

• April 2012: I have been invited to lecture at the International Computer Vision Summer School ICVSS 2012 this July in Sicily.

• April 2012: Opening (1 PhD studentship) - Starting Sept 2012. I am looking for an outstanding candidate to work on dense non-rigid and articulated 3D reconstruction from monocular video. If you are interested please email me directly.

• December 2011: New papers accepted to IJCV and PAMI.

• November 2011: Slides from the Tutorial on Non-Rigid Registration and Reconstruction I gave at ICCV 2011 together with Adrien Bartoli and Alessio Del Bue.

• July 2011: New paper accepted to ICCV 2011.PDF CODE
  Automated Articulated Structure and 3D Shape Recovery from Point Correspondences Joao Fayad, Chris Russell and Lourdes Agapito, in IEEE International Conference in Computer Vision (ICCV 2011 Barcelona, Spain, November 2011.

• March 2011: New paper accepted to CVPR 2011.PDF CODE
  Energy Based Multiple Model Fitting for Non-Rigid SFM Chris Russell, Joao Fayad and Lourdes Agapito, in IEEE Conference in Computer Vision and Pattern Recognition (CVPR 2011 Colorado Springs, USA, June 2011.

RESEARCH TEAM

Postdocs

• Chris Russell :(started October 2010)
  
• Anastasios Roussos : (started October 2010). At Imperial College since November 2014.
  
• Sara Vicente : (started November 2011). At Anthropics since January 2014.

Current PhD Students

• Nikolaos Pitelis : Learning deformable 3D models. (started October 2010)
  
• Rui Yu : Learning dense 3D models from video. (started October 2012)
  

Former PhD Students

• Ravi Garg : Dense 3D reconstruction of non-rigid structure. (2009-2013). Now a postdoc at The Australian Centre for Visual Technologies (ACVT).
  
• Joao Fayad : Piecewise models for 3D reconstruction of non-rigid objects. (2008-2012). Now a JSPS Fellow at the National Institute for Informatics (Japan).
  
• Marco Paladini
  
• Alessio Del Bue

CONTACT