Difference between revisions of "Main Page"

From cvss
 
(24 intermediate revisions by 3 users not shown)
Line 20: Line 20:
 
* Topics may include current research, past research, general topic presentations, paper summaries and critiques, or anything else beneficial to the computer vision graduate student community.
 
* Topics may include current research, past research, general topic presentations, paper summaries and critiques, or anything else beneficial to the computer vision graduate student community.
  
−
==Schedule Spring 2015==
+
==Schedule Fall 2015==
  
 
All talks take place on Thursdays at 3:30pm in AVW 3450.
 
All talks take place on Thursdays at 3:30pm in AVW 3450.
Line 30: Line 30:
 
! Title
 
! Title
 
|-
 
|-
−
| February 19
+
| December 3
−
| Bharat Singh
+
| Angjoo Kanazawa
−
| PSPGC: Part-Based Seeds for Parametric Graph-Cuts
+
| Learning 3D Deformation of Animals from 2D Images
 
|-
 
|-
−
| February 26
+
| December 10
−
| Jingjing Zheng
+
| Xintong Han
−
| Submodular Attribute Selection for Action Recognition in Video
+
| Automated Event Retrieval using Web Trained Detectors
−
|-
 
−
| March 5
 
−
| ''Snow Break''
 
−
|  
 
−
|-
 
−
| March 13
 
−
| Yezhou Yang
 
−
| Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision and Robotics
 
−
|-
 
−
| March 20
 
−
| ''Spring Break, no meeting''
 
−
|
 
−
|-
 
−
| March 27
 
−
| Sravanthi and Varun Manjunatha
 
−
| SHOE: Supervised Hashing with Output Embeddings
 
−
|-
 
−
| April 3
 
−
| Bahadir Ozdemir
 
−
| A Probabilistic Framework for Multimodal Retrieval using Integrative Indian Buffet Process
 
−
|-
 
−
| April 10
 
−
| Ching-Hui Chen
 
−
| Matrix Completion for Resolving Label Ambiguity
 
−
|-
 
−
| April 17
 
−
| ''ICCV deadline, no meeting''
 
−
|
 
−
|-
 
−
| April 24
 
−
| ''post ICCV deadline, no meeting''
 
−
|
 
−
|-
 
−
| May 1
 
−
| Joe Ng
 
−
| Beyond Short Snippets: Deep Networks for Video Classification
 
−
|-
 
−
| May 8
 
−
| Ching Lik Teo
 
−
| Fast 2D Border Ownership Assignment
 
−
|-
 
−
| May 15
 
−
| ''Final Exam, no meeting''
 
−
|
 
 
|}
 
|}
  
 
==Talk Abstracts Spring 2015==
 
==Talk Abstracts Spring 2015==
  
−
===PSPGC: Part-Based Seeds for Parametric Graph-Cuts===
 
−
Speaker: [http://www.cs.umd.edu/~bharat/ Bharat Singh] -- Date: February 19, 2015
 
−
 
−
Abstract: PSPGC is a detection-based parametric graph-cut method for accurate image segmentation. Experiments show that seed positioning plays an important role in graph-cut based methods, so, we propose three seed generation strategies which incorporate information about location and color of object parts, along with size and shape. Combined with low-level regular grid seeds, PSPGC can leverage both low-level and high-level cues about objects present in the image. Multiple-parametric graph-cuts using these seeding strategies are solved to obtain a pool of segments, which have a high rate of producing the ground truth segments. Experiments on the challenging PASCAL2010 and 2012 segmentation datasets show that the accuracy of the segmentation hypotheses generated by PSPGC outperforms other state-of-the-art methods when measured by three different metrics(average overlap, recall and covering) by up to 3.5%. We also obtain the best average overlap score in 15 out of 20 categories on PASCAL2010. Further, we provide a quantitative evaluation of the efficacy of each seed generation strategy introduced.
 
−
 
−
===Submodular Attribute Selection for Action Recognition in Video===
 
−
Speaker: [https://sites.google.com/site/jingjingzhengumd/ Jingjing Zheng] -- Date: February 26, 2015
 
−
 
−
Abstract: We present an approach to jointly learn a set of view-specific dictionaries and a common dictionary for cross-view action recognition. The set of  view-specific dictionaries is learned for specific views while the common dictionary is shared across different views. Our approach represents videos in each view using  both the corresponding view-specific dictionary and the common dictionary. More importantly, it encourages the set of videos taken from different views of the same action to have similar sparse representations. In this way, we can align view-specific features in the sparse feature spaces spanned by the view-specific dictionary set and transfer the view-shared features in the sparse feature space spanned by the common dictionary. Meanwhile, the incoherence between the common dictionary and the view-specific dictionary set enables us to exploit the discrimination information encoded in view-specific features and view-shared features separately. In addition, the learned common dictionary not only has the capability to represent actions from  unseen views, but also makes our approach effective in a semi-supervised setting where no correspondence videos exist and only a few labels exist in the target view. Extensive experiments using the multi-view IXMAS dataset demonstrate that our approach outperforms many recent approaches for cross-view action recognition.
 
−
 
−
===Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision and Robotics===
 
−
Speaker: [http://www.umiacs.umd.edu/~yzyang/ Yezhou Yang] -- Date: March 13, 2015
 
−
 
−
Abstract: Our ability to interpret other people's actions hinges crucially on predictions about their intentionality. The grasp type provides crucial information about human action. However, recognizing the grasp type from unconstrained scenes is challenging because of the large variations in appearance, occlusions and  geometric distortions. In this paper, first we present a convolutional neural network to classify functional hand grasp types. Experiments on a public static scene hand data set validate good performance of the presented method. Then we present two applications utilizing grasp type classification: (a) inference of human action intention and (b) fine level manipulation action segmentation.
 
−
Experiments on both tasks demonstrate the usefulness of grasp type as a cognitive feature for computer vision. Furthermore, we will present a system that learns manipulation action plans by processing Youtube cooking instructional videos with the grasp type feature. Its goal is to robustly generate the  sequence of atomic actions of seen longer actions in video in order to acquire knowledge for robots, and further guide it to execute the task.
 
−
 
−
Related Papers:
 
−
* [http://www.umiacs.umd.edu/~yzyang/paper/CVPR2015Grasp_draft.pdf Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision (To appear in CVPR'15)]
 
−
* [http://www.umiacs.umd.edu/~yzyang/paper/YouCookMani_CameraReady.pdf Robot Learning Manipulation Action Plans by “Watching” Unconstrained Videos from the World Wide Web (AAAI'15)]
 
−
* [http://www.umiacs.umd.edu/~yzyang/paper/VSS_action_intention.pdf Does the grasp type reveal action intention? (To appear in VSS'15)]
 
−
 
−
===SHOE: Supervised Hashing with Output Embeddings===
 
−
Speaker: Sravanthi Bondugula and [https://sites.google.com/site/varunmanjunatha/ Varun Manjunatha] -- Date: March 27, 2015
 
−
 
−
Abstract: We present a supervised binary encoding scheme for image retrieval that learns projections by taking into account similarity between classes obtained from output embeddings. Our motivation is that binary hash codes learned in this way improve both the visual quality of retrieval results and existing supervised hashing schemes. We employ a sequential greedy optimization that learns relationship aware projections by minimizing the difference between inner products of binary codes and output embedding vectors. We develop a joint optimization framework to learn projections which improve the accuracy of supervised hashing over the current state of the art with respect to standard and sibling evaluation metrics. We further boost performance by applying the supervised dimensionality reduction technique on kernelized input CNN features. Experiments are performed on three datasets: CUB-2011, SUN-Attribute and ImageNet ILSVRC 2010. As a by-product of our method, we show that using a simple k-nn pooling classifier with our discriminative codes improves over the complex classification models on fine grained datasets like CUB and offer an impressive compression ratio of 1024 on CNN features.
 
−
 
−
Related paper: [http://arxiv.org/abs/1502.00030 SHOE]
 
−
 
−
===A Probabilistic Framework for Multimodal Retrieval using Integrative Indian Buffet Process===
 
−
Speaker: [http://www.cs.umd.edu/~ozdemir/ Bahadir Ozdemir]
 
−
 
−
Abstract:
 
−
Integrating information from multiple input sources is critical to achieve several key tasks in machine learning. Discovering hidden common causes that explain the dependency among modalities contributes towards enhancing performance in these tasks compared to single-view approaches. We propose a multimodal retrieval procedure based on latent feature models. The procedure consists of a nonparametric Bayesian framework for learning underlying semantically meaningful abstract features in a multimodal dataset, a probabilistic retrieval model that allows cross-modal queries and an extension model for relevance feedback. Experiments on two multimodal datasets, PASCAL-Sentence and SUN-Attribute, demonstrate the effectiveness of the proposed retrieval procedure in comparison to the state-of-the-art algorithms for learning binary codes.
 
−
 
−
Related paper:
 
−
[http://www.cs.umd.edu/~ozdemir/papers/nips14_iibp.pdf A Probabilistic Framework for Multimodal Retrieval using Integrative Indian Buffet Process, NIPS 2014]
 
−
 
−
===Matrix Completion for Resolving Label Ambiguity===
 
−
Speaker: Ching-Hui Chen
 
−
 
−
Abstract: In real applications, data is not always explicitly-labeled. For instance, label ambiguity exists when we associate two persons appearing in a news photo with two names provided in the caption. We propose a matrix completion-based method for resolving ambiguity to predict the actual labels from the ambiguously labeled instances, and a standard supervised classifier can learn from the disambiguated labels to classify new data. We further generalize the method to handle the labeling constraints between instances when such prior knowledge is available. Compared to the state of the arts, our proposed framework achieves 2.9% improvement on the labeling accuracy of the Lost dataset and comparable performance on the Labeled Yahoo! News dataset.
 
  
−
Related paper:
+
===Learning 3D Deformation of Animals from 2D Images===
−
Matrix Completion for Resolving Label Ambiguity, To appear in CVPR 2015
+
Speaker: [http://www.umiacs.umd.edu/~kanazawa/ Angjoo Kanazawa] -- Date: December 3, 2015
  
−
===Beyond Short Snippets: Deep Networks for Video Classification===
+
Abstract: Understanding how an animal can deform and articulate is essential for a realistic modification of its 3D model. In this paper, we show that such information can be learned from user-clicked 2D images and a template 3D model of the target animal. We present a volumetric deformation framework that produces a set of new 3D models by deforming a template 3D model according to a set of user-clicked images. Our framework is based on a novel locally-bounded deformation energy, where every local region has its own stiffness value that bounds how much distortion is allowed at that location. We jointly learn the local stiffness bounds as we deform the template 3D mesh to match each user-clicked image. We show that this seemingly complex task can be solved as a sequence of convex optimization problems. We demonstrate the effectiveness of our approach on cats and horses, which are highly deformable and articulated animals. Our framework produces new 3D models of animals that are significantly more plausible than methods without learned stiffness.
−
Speaker: Joe Ng
 
  
−
Abstract: Convolutional neural networks (CNNs) have been extensively applied for image recognition problems giving state-of-the-art results on recognition, detection, segmentation and retrieval. In this work we propose and evaluate several deep neural network architectures to combine image information across a video over longer time periods than previously attempted. We propose two methods capable of handling full length videos. The first method explores various convolutional temporal feature pooling architectures, examining the various design choices which need to be made when adapting a CNN for this task. The second proposed method explicitly models the video as an ordered sequence of frames. For this purpose we employ a recurrent neural network that uses Long Short-Term Memory (LSTM) cells which are connected to the output of the underlying CNN. Our best networks exhibit significant performance improvements over previously published results on the Sports 1 million dataset (73.1% vs. 60.9%) and the UCF-101 datasets with (88.6% vs. 87.9%) and without additional optical flow information (82.6% vs. 72.8%).
+
Link: [http://arxiv.org/pdf/1507.07646v1.pdf paper]
  
−
Link: [http://arxiv.org/abs/1503.08909 Preliminary Version] (To appear in CVPR 2015)
+
===Automated Event Retrieval using Web Trained Detectors===
  
−
===Fast 2D Border Ownership Assignment===
+
Speaker: [http://www.umiacs.umd.edu/~xintong/ Xintong Han] -- Date: December 10, 2015
−
Speaker: Ching-Lik Teo
 
  
−
A method for efficient border ownership assignment in 2D images is proposed. Leveraging on recent advances using Structured Random Forests (SRF) for boundary detection, we impose a novel border ownership structure that detects both boundaries and border ownership at the
+
Abstract: Complex event retrieval is a challenging research problem, especially when no training videos are available. An alternative to collecting training videos is to train a large semantic concept bank a priori. Given a text description of an event, event retrieval is performed by selecting concepts linguistically related to the event description and fusing the concept responses on unseen videos. However, defining an exhaustive concept lexicon and pre-training it requires vast computational resources. Therefore, recent approaches automate concept discovery and training by leveraging large amounts of weakly annotated web data. Compact visually salient concepts are automatically obtained by the use of concept pairs or, more generally, n-grams. However, not all visually salient n-grams are necessarily useful for an event query - some combinations of concepts may be visually compact but irrelevant--and this drastically affects performance. We propose an event retrieval algorithm that constructs pairs of automatically discovered concepts and then prunes those concepts that are unlikely to be helpful for retrieval. Pruning depends both on the query and on the specific video instance being evaluated. Our approach also addresses calibration and domain adaptation issues that arise when applying concept detectors to unseen videos. We demonstrate large improvements over other vision based systems on the TRECVID MED 13 dataset.
−
same time. Key to this work are features that predict ownership cues from 2D images. To this end, we use several different local cues: shape, spectral properties of boundary patches, and semi-global grouping cues that are indicative of perceived depth. For shape, we use HoG-like descriptors that encode local curvature (convexity and concavity). For spectral properties, such as extremal edges (EE), we first learn an orthonormal basis spanned by the top K eigenvectors via PCA over common types of contour tokens from which we reproject the patches to extract the most important spectral features. For grouping, we introduce a novel mid-level descriptor that captures patterns near edges and indicates ownership information of the boundary. Experimental results over a subset of the Berkeley Segmentation Dataset (BSDS) and the NYU Depth V2 dataset show that our method’s performance exceeds current state of the art multi-stage approaches that use more complex features.
 
  
−
Related paper: [http://www.umiacs.umd.edu/%7Ecteo/public-shared/CVPR15_BorderOwnership_final.pdf PDF] C.L. Teo, C. Fermüller, Y. Aloimonos. Fast 2D Border Ownership Assignment. IEEE Conf. on
+
Link: [http://arxiv.org/pdf/1509.07845v1.pdf paper]
−
Computer Vision and Pattern Recognition (CVPR), to appear, 2015.
 
  
 
==Past Semesters==
 
==Past Semesters==
 +
* [[Cvss:Spring2015| Spring 2015]]
 
* [[cvss fall2014|Fall 2014]]
 
* [[cvss fall2014|Fall 2014]]
 
* [[cvss_spring2014|Spring 2014]]
 
* [[cvss_spring2014|Spring 2014]]
Line 159: Line 71:
 
==Funded By==
 
==Funded By==
 
* Computer Vision Faculty
 
* Computer Vision Faculty
−
* '''[http://www.northropgrumman.com/ Northrop Grumman]'''
+
<!-- * '''[http://www.northropgrumman.com/ Northrop Grumman]''' -->
  
 
==Current Seminar Series Coordinators==
 
==Current Seminar Series Coordinators==
Line 167: Line 79:
 
{| cellpadding="1"
 
{| cellpadding="1"
 
|-
 
|-
−
| [http://www.umiacs.umd.edu/~jhchoi/ Jonghyun Choi], jhchoi@
+
| [http://sites.google.com/site/austinomyers/ Austin Myers], amyers@
−
| (student of [http://www.umiacs.umd.edu/~lsd/ Professor Larry Davis])
 
−
|-
 
−
| [https://sites.google.com/site/austinomyers/ Austin Myers], amyers@
 
 
| (student of [http://www.cfar.umd.edu/~yiannis/ Professor Yiannis Aloimonos])
 
| (student of [http://www.cfar.umd.edu/~yiannis/ Professor Yiannis Aloimonos])
 
|-
 
|-
Line 176: Line 85:
 
| (student of [http://cs.umd.edu/~djacobs/ Professor David Jacobs])
 
| (student of [http://cs.umd.edu/~djacobs/ Professor David Jacobs])
 
|-
 
|-
−
| Ching-Hui Chen, ching@
+
| [http://sites.google.com/site/yechengxi/ Chenxi Ye] cxy@
−
| (student of [http://www.umiacs.umd.edu/~rama/ Professor Rama Chellappa])
+
| (student of [http://www.cfar.umd.edu/~yiannis/ Professor Yiannis Aloimonos])
 +
|-
 +
| [http://www.umiacs.umd.edu/~xintong/ Xintong Han], xintong@
 +
| (student of [http://www.umiacs.umd.edu/~lsd/ Professor Larry Davis])
 +
|-
 +
| [http://www.cs.umd.edu/~bharat/ Bharat Singh], bharat@
 +
| (student of [http://www.umiacs.umd.edu/~lsd/ Professor Larry Davis])
 +
|-
 +
| [http://bcsiriuschen.github.io/ Bor-Chun (Sirius) Chen], sirius@
 +
| (student of [http://www.umiacs.umd.edu/~lsd/ Professor Larry Davis])
 
|}
 
|}
  
 
Gone but not forgotten.
 
Gone but not forgotten.
−
 
 
{| cellpadding="1"
 
{| cellpadding="1"
 +
|-
 +
| [http://www.umiacs.umd.edu/~jhchoi/ Jonghyun Choi], jhchoi@
 +
| (student of [http://www.umiacs.umd.edu/~lsd/ Professor Larry Davis])
 +
|-
 +
| Ching-Hui Chen, ching@
 +
| (student of [http://www.umiacs.umd.edu/~rama/ Professor Rama Chellappa])
 +
|
 
|-
 
|-
 
| [http://ravitejav.weebly.com/ Raviteja Vemulapalli], raviteja @
 
| [http://ravitejav.weebly.com/ Raviteja Vemulapalli], raviteja @

Latest revision as of 23:40, 3 December 2015

Computer Vision Student Seminars

The Computer Vision Student Seminars at the University of Maryland College Park are a student-run series of talks given by current graduate students for current graduate students.

To receive regular information about the Computer Vision Student Seminars, subscribe to our mailing list or our talks list.

Description[edit]

The purpose of these talks is to:

  • Encourage interaction between computer vision students;
  • Provide an opportunity for computer vision students to be aware of and possibly get involved in the research their peers are conducting;
  • Provide an opportunity for computer vision students to receive feedback on their current research;
  • Provide speaking opportunities for computer vision students.

The guidelines for the format are:

  • An hour-long weekly meeting, consisting of one 20-40 minute talk followed by discussion and food.
  • The talks are meant to be casual and discussion is encouraged.
  • Topics may include current research, past research, general topic presentations, paper summaries and critiques, or anything else beneficial to the computer vision graduate student community.

Schedule Fall 2015[edit]

All talks take place on Thursdays at 3:30pm in AVW 3450.

Date Speaker Title
December 3 Angjoo Kanazawa Learning 3D Deformation of Animals from 2D Images
December 10 Xintong Han Automated Event Retrieval using Web Trained Detectors

Talk Abstracts Spring 2015[edit]

Learning 3D Deformation of Animals from 2D Images[edit]

Speaker: Angjoo Kanazawa -- Date: December 3, 2015

Abstract: Understanding how an animal can deform and articulate is essential for a realistic modification of its 3D model. In this paper, we show that such information can be learned from user-clicked 2D images and a template 3D model of the target animal. We present a volumetric deformation framework that produces a set of new 3D models by deforming a template 3D model according to a set of user-clicked images. Our framework is based on a novel locally-bounded deformation energy, where every local region has its own stiffness value that bounds how much distortion is allowed at that location. We jointly learn the local stiffness bounds as we deform the template 3D mesh to match each user-clicked image. We show that this seemingly complex task can be solved as a sequence of convex optimization problems. We demonstrate the effectiveness of our approach on cats and horses, which are highly deformable and articulated animals. Our framework produces new 3D models of animals that are significantly more plausible than methods without learned stiffness.

Link: paper

Automated Event Retrieval using Web Trained Detectors[edit]

Speaker: Xintong Han -- Date: December 10, 2015

Abstract: Complex event retrieval is a challenging research problem, especially when no training videos are available. An alternative to collecting training videos is to train a large semantic concept bank a priori. Given a text description of an event, event retrieval is performed by selecting concepts linguistically related to the event description and fusing the concept responses on unseen videos. However, defining an exhaustive concept lexicon and pre-training it requires vast computational resources. Therefore, recent approaches automate concept discovery and training by leveraging large amounts of weakly annotated web data. Compact visually salient concepts are automatically obtained by the use of concept pairs or, more generally, n-grams. However, not all visually salient n-grams are necessarily useful for an event query - some combinations of concepts may be visually compact but irrelevant--and this drastically affects performance. We propose an event retrieval algorithm that constructs pairs of automatically discovered concepts and then prunes those concepts that are unlikely to be helpful for retrieval. Pruning depends both on the query and on the specific video instance being evaluated. Our approach also addresses calibration and domain adaptation issues that arise when applying concept detectors to unseen videos. We demonstrate large improvements over other vision based systems on the TRECVID MED 13 dataset.

Link: paper

Past Semesters[edit]

Funded By[edit]

  • Computer Vision Faculty

Current Seminar Series Coordinators[edit]

Emails are at umiacs.umd.edu.

Austin Myers, amyers@ (student of Professor Yiannis Aloimonos)
Angjoo Kanazawa, kanazawa@ (student of Professor David Jacobs)
Chenxi Ye cxy@ (student of Professor Yiannis Aloimonos)
Xintong Han, xintong@ (student of Professor Larry Davis)
Bharat Singh, bharat@ (student of Professor Larry Davis)
Bor-Chun (Sirius) Chen, sirius@ (student of Professor Larry Davis)

Gone but not forgotten.

Jonghyun Choi, jhchoi@ (student of Professor Larry Davis)
Ching-Hui Chen, ching@ (student of Professor Rama Chellappa)
Raviteja Vemulapalli, raviteja @ (student of Professor Rama Chellappa)
Sameh Khamis
Ejaz Ahmed
Anne Jorstad now at EPFL
Jie Ni now at Sony
Sima Taheri
Ching Lik Teo

Web Accessibility

Retrieved from "https://wiki.cs.umd.edu/cvss/w/index.php?title=Main_Page&oldid=1868"