Hauptseminar/Master Seminar:
3D Computer Vision

Prof. Nassir Navab, Federico Tombari, Vasileios Belagiannis, Felix Achilles, Wadim Kehl, Benjamin Busam

Type: Advanced Seminar Course Module IN8901
Type: Master Seminar Module IN2107
Type: Seminar Module IN0014
SWS: 2+0
ECTS: 4 Credits
Location: MI 03.13.010
Time: Tuesday 14:00-16:00
Course Language: English

Announcements

• [3.11.2014] The list of papers has been published
• [16.10.2014] If interested, please register here to the course
• [16.10.2014] Initial meeting will be held on Tuesday, 04.11.2014, in MI 03.13.010 from 15:00 to 16:00.
  
• [16.10.2014] The list of papers will be online one day before the initial meeting.

Introduction

• The Seminar Course concerns recent advances in the field of 3D Computer Vision. The Seminar will propose a list of recent scientific articles related to the main current research topics of 3D Computer Vision, such as 3D keypoint detection, 3D features, point cloud registration, 3D object recognition and classification, SLAM and stereo vision.

Registration

• Registration can be performed through TUM online.
• Write a short email indicating your interest on the course to tombari@in.tum.de

Requirements

In this Master Seminar (formerly Hauptseminar), each student is asked to select one paper from a list. In order to successfully complete the seminar, participants have to fulfill these requirements:

• Presentation: The selected paper is presented to the other participants (25 minutes presentation 15 minutes questions). Use the CAMP templates for PowerPoint CAMPmaster.pot, or Latex: CAMP-latex-template.zip.
• Written Handout: A document of maximum 8 pages is written and submitted one week after the presentation. Please download and use the llncs2e.zip Latex template.
• Attendance: Participants have to participate actively in all seminar sessions.

The students are required to attend each seminar presentation which will be held during this course. Each presentation is followed by a discussion and everyone is encouraged to actively participate. The handout must include all references used and must be written completely in your own words. Copy and paste will not be tolerated. Both handout and presentation have to be done in English.

List of Topics and Material

The list of papers is the following:

• (features, detection) J. Sun, M. Ovsjanikov, L. Guibas, “A Concise and Provably Informative Multi-Scale Signature Based on Heat Diffusion”, Comput. Graph. Forum, Vol. 28:5, pages 1383-1392, 2009
• (features, detection) F. Tombari, S. Salti, L. Di Stefano, “Unique signatures of Histograms for local surface description”, ECCV 2010
• (3D/RGBD object recognition and classification) M. Blum, J.T. Springenberg, J. Wulfing, M. Riedmiller, “A Learned Feature Descriptor for Object Recognition in RGB-D Data”, ICRA 2012
• (3D/RGBD object recognition and classification) R. Socher, B. Huval, B. Bhat, C.D. Manning, A.Y. Ng, “Convolutional-Recursive Deep Learning for 3D Object Classification”, NIPS 2012
• (3D semantic segmentation) D. Munoz, J. A. Bagnell, N. Vandapel, M. Hebert , “Contextual Classification with Functional Max-Margin Markov Networks”, CVPR 2009
• (3D reconstruction) Newcombe et al., “KinectFusion: Real-Time Dense Surface Mapping and Tracking”, ISMAR 2011
• (point cloud registration) Segal, Haehnel, Thrun, “Generalized-ICP”, RSS 2005
• (stereo vision) C. Rhemann, A. Hosni, M. Bleyer, C. Rother, M. Gelautz, “Fast cost-volume filtering for visual correspondence and beyond”, CVPR 2011
• (3D human pose estimation) J. Shotton et al., “Real-time human pose recognition in parts from single depth images”, CVPR 2011
• (3D human pose estimation) S. Li, A. Chan, “3D Human Pose Estimation from Monocular Images with Deep Convolutional Neural Network”, ACCV 2014

The slides for the introductory meeting are available here.

Schedule

This is the final schedule for the course. Each registered student is required to attend all presentations. All lectures take place in Seminarraum 03.13.010.

Date	Time	Topic	Speaker	Supervision	Material
2.12	2:00-2:40 pm	KinectFusion?: Real-Time Dense Surface Mapping and Tracking	Fabian Manhardt	Wadim Kehl
2.12	2:40-3:20 pm	Real-time human pose recognition in parts from single depth images	Ruchik Yajnik	Vasileios Belagionnis
2.12	3:20-4:00 pm	A Concise and Provably Informative Multi-Scale Signature Based on Heat Diffusion	Maximilian Jakasovic	Felix Achilles
9.12	2:00-2:40 pm	3D Human Pose Estimation from Monocular Images with Deep Convolutional Neural Network	Simon Trendel	Vasileios Belagionnis
9.12	2:40-3:20 pm	A Learned Feature Descriptor for Object Recognition in RGB-D Data	Nasir Ashfaq	Vasileios Belagionnis
9.12	3:20-4:00 pm	Convolutional-Recursive Deep Learning for 3D Object Classification	Stefan Matthes	Wadim Kehl