The PASCAL Visual Object Classes Challenge Workshop 2012

12th October 2012, ECCV 2012, Florence, Italy

Mark Everingham

It is with great sadness that we report that Mark Everingham has died. Mark was the key member of the VOC project, and it would have been impossible without his selfless contributions. This workshop is dedicated to Mark's memory. A tribute web page has been set up.

Introduction

The PASCAL Visual Object Classes (VOC) Challenge has been run annually since 2005. The aim of the challenge is to evaluate the state-of-the-art in object recognition methods, with competitions on object classification, detection and segmentation, large scale recognition, action recognition in still images and person layout estimation. See the challenge website for more details. An annual workshop has been held in association with one of the main vision conferences (ECCV/ICCV) to present and discuss results of the challenge and the methods employed by participants.

The VOC challenge will be run for the final time in 2012. The challenge workshop this year will therefore take a different format than in previous years. There will be two components to the workshop:

The workshop will present methods and results of the VOC2012 challenge and the ImageNet large scale visual recognition challenge. As in previous years, selected speakers will be invited from participants in the challenge, with the emphasis firmly on novelty of submitted methods, rather than simply on the quantitative performance.
The workshop will look back over the eight years of the challenge and provide an opportunity for discussion of lessons learnt from the running of the challenge, progress in the object recognition area over this time, and emerging directions for future work. This part of the workshop will take the form of invited keynote talks, from some of the most influential researchers in the object recognition community. A list of speakers is provided below.

Venue

The workshop will be held on 12th October 2012, in association with ECCV 2012. Further details of the venue can be found on the ECCV website.

Invited Speakers

Derek Hoiem (UIUC)
Jitendra Malik (UC Berkeley)
Deva Ramanan (UC Irvine)

Prizes

The following prizes were announced at the workshop:

Classification
Winner:	NUSPSL_CTX_GPM_SCM Dong Jian, Chen Qiang, Song Zheng, Pan Yan, Xia Wei, Yan Shuicheng, Hua Yang, Huang Zhongyang, Shen Shengmei National University of Singapore; Panasonic Singapore Laboratories; Sun Yat-sen University
Honourable Mention:	CVC_UVA_UNITN_BOW_FK_COLORDET_SP Fahad Khan, Jan van Gemert, Camp Davesa, Jasper Uijlings , Albert Gordo, Sezer Karaoglu, Koen van de Sande, Pep Gonfaus, Rao Muhammad Anwer, Joost van de Weijer, Cees Snoek, Ramon Baldrich, Nicu Sebe, Theo gevers Computer Vision Barcelona; University of Amsterdam; University of Trento
Detection
Winner:	UVA_HYBRID_CODING_APE Koen E. A. van de Sande, Jasper R. R. Uijlings, Cees G. M. Snoek, Arnold W. M. Smeulders University of Amsterdam
Honourable Mention:	OXFORD_DPM_MKL Ross Girshick, Andrea Vedaldi, Karen Simonyan University of Oxford
Segmentation
Winner (Comp 5):	NUS_DET_SPR_GC_SP Wei Xia, Csaba Domokos, Jian Dong, Loong Fah Cheong, Shuicheng Yan; Zhongyang Huang, Shengmei Shen National University of Singapore; Panasonic Singapore Laboratories
Winner (Comp 6):	BONN_O2PCPMC_FGT_SEGM Joƒo Carreira, Adrian Ion, Fuxin Li, Cristian Sminchisescu University of Bonn
Action Classification
Winner (Comp 9):	STANFORD_RF_MULTFEAT_SVM Aditya Khosla, Rui Zhang, Bangpeng Yao, Li Fei-Fei Stanford University, MIT
Winner (Comp 10):	OXFORD_ALIGNED_BODYPARTS Minh Hoai, Lubor Ladicky, Andrew Zisserman University of Oxford

Programme

Part I: VOC 2012 - methods, results and analysis

09:15 - 10.05	Introduction, overview and results of the classification challenge Andrew Zisserman [PDF]
	Generalized Hierarchical Matching for Sub-category Aware Object Classification Shuicheng YAN, Jian DONG, Qiang CHEN, Zheng SONG, Yan PAN, Wei XIA; Zhongyang HUANG , Yang HUA, Shengmei SHEN National University of Singapore; Panasonic Singapore Laboratories [PDF] NB: This is unpublished work. Please contact the authors if you plan to make use of any of the ideas presented.
10:05 - 10:40	Overview and results of the detection challenge Chris Williams [PDF]
	Hybrid Coding for Selective Search Koen van de Sande , Jasper Uijlings, Cees Snoek, Arnold Smeulders University of Amsterdam, Netherlands [PDF] NB: This is unpublished work. Please contact the authors if you plan to make use of any of the ideas presented.
10:40 - 11:00	Break
11:00 - 11:20	Detection challenge concluded
11:20 - 11:40	Overview and results of the segmentation challenge, and action taster Luc Van Gool [PDF]

Part II: VOC 2005-2012: the VOC years and legacy

11:40 - 12:00	Introduction: history and analysis Chris Williams [PDF]
	One minute silence for Mark Everingham
12:00 - 12:30	"Deformable Part Models: Some thoughts on why they work and what next" Prof. Deva Ramanan UC Irvine [PDF] [Videos]
12:30 - 13:00	"What Next? Progress and Pressing Challenges in Object Recognition" Prof Derek Hoiem UIUC [PDF]
13:00 - 14:30	Lunch
14:30 - 15:00	"The Three R's of Computer Vision: Recognition, Reconstruction and Reorganization" Prof Jitendra Malik UC Berkeley [PDF]
15:00 - 15:30	Conclusions and close The VOC organizers [PDF]

15:30 - 18:15

ILSVRC2012 challenge workshop
More information available at the ILSVRC2012 website

Organizers

Mark Everingham (University of Leeds)
Luc van Gool (ETHZ, Zurich), vangool@vision.ee.ethz.ch
Chris Williams (University of Edinburgh)
John Winn (Microsoft Research Cambridge)
Andrew Zisserman (University of Oxford)

Acknowledgements

The organizers would particularly like to thank Ali Eslami and Yusuf Aytar for their substantial efforts in preparing the results, slides and web pages for the workshop.

Support

The preparation and running of this challenge is supported by the EU-funded PASCAL2 Network of Excellence on Pattern Analysis, Statistical Modelling and Computational Learning.