Sorry, you need to enable JavaScript to visit this website.

Image/Video Storage, Retrieval

Multi-view deep metric learning for image classification

Paper Details

Authors:
Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju
Submitted On:
14 September 2017 - 10:53am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

poster_MVDML.pdf

(279)

Subscribe

[1] Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju, "Multi-view deep metric learning for image classification", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/2034. Accessed: Sep. 20, 2020.
@article{2034-17,
url = {http://sigport.org/2034},
author = {Dewei Li; Jingjing Tang; Yingjie Tian; Xuchan Ju },
publisher = {IEEE SigPort},
title = {Multi-view deep metric learning for image classification},
year = {2017} }
TY - EJOUR
T1 - Multi-view deep metric learning for image classification
AU - Dewei Li; Jingjing Tang; Yingjie Tian; Xuchan Ju
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/2034
ER -
Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju. (2017). Multi-view deep metric learning for image classification. IEEE SigPort. http://sigport.org/2034
Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju, 2017. Multi-view deep metric learning for image classification. Available at: http://sigport.org/2034.
Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju. (2017). "Multi-view deep metric learning for image classification." Web.
1. Dewei Li, Jingjing Tang, Yingjie Tian, Xuchan Ju. Multi-view deep metric learning for image classification [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/2034

LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL


Matrix factorization based hashing has been very effective in addressing the cross-modal retrieval task. In this work, we propose a novel supervised hashing approach utilizing the concepts of matrix factorization which can seamlessly incorporate the label information. In the proposed approach, the latent factors for each individual modality are generated, which are then converted to the more discriminative label space using modality specific linear transformations.

Paper Details

Authors:
Devraj Mandal, Soma Biswas
Submitted On:
14 September 2017 - 8:30am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

The poster of the paper.

(591)

Subscribe

[1] Devraj Mandal, Soma Biswas, "LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/2023. Accessed: Sep. 20, 2020.
@article{2023-17,
url = {http://sigport.org/2023},
author = {Devraj Mandal; Soma Biswas },
publisher = {IEEE SigPort},
title = {LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL},
year = {2017} }
TY - EJOUR
T1 - LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL
AU - Devraj Mandal; Soma Biswas
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/2023
ER -
Devraj Mandal, Soma Biswas. (2017). LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL. IEEE SigPort. http://sigport.org/2023
Devraj Mandal, Soma Biswas, 2017. LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL. Available at: http://sigport.org/2023.
Devraj Mandal, Soma Biswas. (2017). "LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL." Web.
1. Devraj Mandal, Soma Biswas. LABEL CONSISTENT MATRIX FACTORIZATION BASED HASHING FOR CROSS-MODAL RETRIEVAL [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/2023

Summarization of Human Activity Videos Using a Salient Dictionary


Video summarization has become more prominent during the last decade, due to the massive amount of available digital video content. A video summarization algorithm is typically fed an input video and expected to extract a set of important key-frames which represent the entire content, convey semantic meaning and are significantly more concise than the original input. The most wide-spread approach relies on video frame clustering and extraction of the frames closest to the cluster centroids as key-frames.

Paper Details

Authors:
Anastasios Tefas, Ioannis Pitas
Submitted On:
13 September 2017 - 11:18am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Key-frame extraction from human activity videos via salient dictionary learning-based video summarization

(353)

Subscribe

[1] Anastasios Tefas, Ioannis Pitas, "Summarization of Human Activity Videos Using a Salient Dictionary", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1976. Accessed: Sep. 20, 2020.
@article{1976-17,
url = {http://sigport.org/1976},
author = {Anastasios Tefas; Ioannis Pitas },
publisher = {IEEE SigPort},
title = {Summarization of Human Activity Videos Using a Salient Dictionary},
year = {2017} }
TY - EJOUR
T1 - Summarization of Human Activity Videos Using a Salient Dictionary
AU - Anastasios Tefas; Ioannis Pitas
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1976
ER -
Anastasios Tefas, Ioannis Pitas. (2017). Summarization of Human Activity Videos Using a Salient Dictionary. IEEE SigPort. http://sigport.org/1976
Anastasios Tefas, Ioannis Pitas, 2017. Summarization of Human Activity Videos Using a Salient Dictionary. Available at: http://sigport.org/1976.
Anastasios Tefas, Ioannis Pitas. (2017). "Summarization of Human Activity Videos Using a Salient Dictionary." Web.
1. Anastasios Tefas, Ioannis Pitas. Summarization of Human Activity Videos Using a Salient Dictionary [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1976

Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics


Video summarization is a timely and rapidly developing research field with broad commercial interest, due to the increasing availability of massive video data. Relevant algorithms face the challenge of needing to achieve a careful balance between summary compactness, enjoyability and content coverage. The specific case of stereoscopic $3$D theatrical films has become more important over the past years, but not received corresponding research attention.

Paper Details

Authors:
Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas
Submitted On:
13 September 2017 - 11:12am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Stereoscopic Movie Summarization Conforming to Narrative Characteristics

(374)

Subscribe

[1] Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas, "Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1975. Accessed: Sep. 20, 2020.
@article{1975-17,
url = {http://sigport.org/1975},
author = {Anastasios Tefas; Nikos Nikolaidis; Ioannis Pitas },
publisher = {IEEE SigPort},
title = {Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics},
year = {2017} }
TY - EJOUR
T1 - Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics
AU - Anastasios Tefas; Nikos Nikolaidis; Ioannis Pitas
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1975
ER -
Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas. (2017). Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics. IEEE SigPort. http://sigport.org/1975
Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas, 2017. Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics. Available at: http://sigport.org/1975.
Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas. (2017). "Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics." Web.
1. Anastasios Tefas, Nikos Nikolaidis, Ioannis Pitas. Multimodal Stereoscopic Movie Summarization Conforming to Narrative Characteristics [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1975

MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING


The multiple types of social media data have abundant information, but learning multi-modal social data is challenging due to data heterogeneity and noise in user-generated data. To address this problem, we propose a multi-view network-based clustering approach that is robust to noise and fully reflects the underlying structure of the comprehensive network. To demonstrate the proposed approach, we experimented with clustering challenging tagged images of landmarks.

Paper Details

Authors:
So Yeon Kim, Kyung-Ah Sohn
Submitted On:
13 September 2017 - 5:22am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICIP2017_poster.pdf

(265)

Subscribe

[1] So Yeon Kim, Kyung-Ah Sohn, "MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1967. Accessed: Sep. 20, 2020.
@article{1967-17,
url = {http://sigport.org/1967},
author = {So Yeon Kim; Kyung-Ah Sohn },
publisher = {IEEE SigPort},
title = {MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING},
year = {2017} }
TY - EJOUR
T1 - MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING
AU - So Yeon Kim; Kyung-Ah Sohn
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1967
ER -
So Yeon Kim, Kyung-Ah Sohn. (2017). MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING. IEEE SigPort. http://sigport.org/1967
So Yeon Kim, Kyung-Ah Sohn, 2017. MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING. Available at: http://sigport.org/1967.
So Yeon Kim, Kyung-Ah Sohn. (2017). "MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING." Web.
1. So Yeon Kim, Kyung-Ah Sohn. MULTI-VIEW NETWORK-BASED SOCIAL-TAGGED LANDMARK IMAGE CLUSTERING [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1967

POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION


For 3D object detection and pose estimation, it is crucial to extract distinctive and representative features of the objects and describe them efficiently. Therefore, a large number of 3D feature descriptors has been developed. Among these, Point Feature Histogram RGB (PFHRGB) has been evaluated as showing the best performance for 3D object and category recognition. However, this descriptor is vulnerable to point density variation and produces many false correspondences accordingly.

Paper Details

Authors:
Submitted On:
12 September 2017 - 12:20pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

icip_poster_suakim.pdf

(237)

Subscribe

[1] , "POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1947. Accessed: Sep. 20, 2020.
@article{1947-17,
url = {http://sigport.org/1947},
author = { },
publisher = {IEEE SigPort},
title = {POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION},
year = {2017} }
TY - EJOUR
T1 - POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION
AU -
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1947
ER -
. (2017). POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION. IEEE SigPort. http://sigport.org/1947
, 2017. POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION. Available at: http://sigport.org/1947.
. (2017). "POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION." Web.
1. . POINT DENSITY-INVARIANT 3D OBJECT DETECTION AND POSE ESTIMATION [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1947

A Pool of Deep Models for Event Recognition


This paper proposes a novel two-stage framework for event recognition in still images. First, for a generic event image, deep features, obtained via different pre-trained models, are fed into an ensemble of classifiers, whose posterior classification probabilities are thereafter fused by means of an order induced scheme, which penalizes the yielded scores according to their confidence in classifying the image at hand, and then averages them. Second, we combine the fusion results with a reverse matching paradigm in order to draw the final output of our proposed pipeline.

Paper Details

Authors:
Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale
Submitted On:
12 September 2017 - 11:01am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICIP_2017_updated.pptx

(309)

Subscribe

[1] Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale, "A Pool of Deep Models for Event Recognition", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1942. Accessed: Sep. 20, 2020.
@article{1942-17,
url = {http://sigport.org/1942},
author = {Kashif Ahmad;Mohamed Lamine Mekhalfi; Nicola Conci; Giulia Boato; Farid Megani; Francesco De Natale },
publisher = {IEEE SigPort},
title = {A Pool of Deep Models for Event Recognition},
year = {2017} }
TY - EJOUR
T1 - A Pool of Deep Models for Event Recognition
AU - Kashif Ahmad;Mohamed Lamine Mekhalfi; Nicola Conci; Giulia Boato; Farid Megani; Francesco De Natale
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1942
ER -
Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. (2017). A Pool of Deep Models for Event Recognition. IEEE SigPort. http://sigport.org/1942
Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale, 2017. A Pool of Deep Models for Event Recognition. Available at: http://sigport.org/1942.
Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. (2017). "A Pool of Deep Models for Event Recognition." Web.
1. Kashif Ahmad,Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. A Pool of Deep Models for Event Recognition [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1942

A Pool of Deep Models for Event Recognition


This paper proposes a novel two-stage framework for event recognition in still images. First, for a generic event image, deep features, obtained via different pre-trained models, are fed into an ensemble of classifiers, whose posterior classification probabilities are thereafter fused by means of an order induced scheme, which penalizes the yielded scores according to their confidence in classifying the image at hand, and then averages them. Second, we combine the fusion results with a reverse matching paradigm in order to draw the final output of our proposed pipeline.

Paper Details

Authors:
Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale
Submitted On:
12 September 2017 - 11:01am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICIP_2017_updated.pptx

(250)

Subscribe

[1] Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale, "A Pool of Deep Models for Event Recognition", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1941. Accessed: Sep. 20, 2020.
@article{1941-17,
url = {http://sigport.org/1941},
author = {Mohamed Lamine Mekhalfi; Nicola Conci; Giulia Boato; Farid Megani; Francesco De Natale },
publisher = {IEEE SigPort},
title = {A Pool of Deep Models for Event Recognition},
year = {2017} }
TY - EJOUR
T1 - A Pool of Deep Models for Event Recognition
AU - Mohamed Lamine Mekhalfi; Nicola Conci; Giulia Boato; Farid Megani; Francesco De Natale
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1941
ER -
Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. (2017). A Pool of Deep Models for Event Recognition. IEEE SigPort. http://sigport.org/1941
Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale, 2017. A Pool of Deep Models for Event Recognition. Available at: http://sigport.org/1941.
Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. (2017). "A Pool of Deep Models for Event Recognition." Web.
1. Mohamed Lamine Mekhalfi, Nicola Conci, Giulia Boato, Farid Megani, Francesco De Natale. A Pool of Deep Models for Event Recognition [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1941

Learning a Cross-Modal Hashing Network for Multimedia Search


In this paper, we propose a cross-modal hashing network (CMHN) method to learn compact binary codes for cross-modality multimedia search. Unlike most existing cross-modal hashing methods which learn a single pair of projections to map each example into a binary vector, we design a deep neural network to learn multiple pairs of hierarchical non-linear transformations, under which the nonlinear characteristics of samples can be well exploited and the modality gap is well reduced.

Paper Details

Authors:
Venice Erin Liong, Jiwen Lu, Yap-Peng Tan
Submitted On:
11 September 2017 - 5:44am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

icip2017_poster_2555.pdf

(273)

Subscribe

[1] Venice Erin Liong, Jiwen Lu, Yap-Peng Tan, "Learning a Cross-Modal Hashing Network for Multimedia Search", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1901. Accessed: Sep. 20, 2020.
@article{1901-17,
url = {http://sigport.org/1901},
author = {Venice Erin Liong; Jiwen Lu; Yap-Peng Tan },
publisher = {IEEE SigPort},
title = {Learning a Cross-Modal Hashing Network for Multimedia Search},
year = {2017} }
TY - EJOUR
T1 - Learning a Cross-Modal Hashing Network for Multimedia Search
AU - Venice Erin Liong; Jiwen Lu; Yap-Peng Tan
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1901
ER -
Venice Erin Liong, Jiwen Lu, Yap-Peng Tan. (2017). Learning a Cross-Modal Hashing Network for Multimedia Search. IEEE SigPort. http://sigport.org/1901
Venice Erin Liong, Jiwen Lu, Yap-Peng Tan, 2017. Learning a Cross-Modal Hashing Network for Multimedia Search. Available at: http://sigport.org/1901.
Venice Erin Liong, Jiwen Lu, Yap-Peng Tan. (2017). "Learning a Cross-Modal Hashing Network for Multimedia Search." Web.
1. Venice Erin Liong, Jiwen Lu, Yap-Peng Tan. Learning a Cross-Modal Hashing Network for Multimedia Search [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1901

SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION


Person re-identification aims to match people across disjoint camera views. It has been reported that Least Square
Semi-Coupled Dictionary Learning (LSSCDL) based samplespecific SVM learning framework has obtained the state of
the art performance. However, the objective function of the LSSCDL, the algorithm of learning the pairs (feature, weight)
dictionaries and the mapping function between feature space and weight space, is non-convex, which usually result in

Paper Details

Authors:
Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang
Submitted On:
7 September 2017 - 11:10am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

weixu_ICIP2017poster.pdf

(275)

Subscribe

[1] Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang, "SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1854. Accessed: Sep. 20, 2020.
@article{1854-17,
url = {http://sigport.org/1854},
author = {Wei Xu; Haoyuan Chi; Lei Zhou; Xiaolin Huang; Jie Yang },
publisher = {IEEE SigPort},
title = {SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION},
year = {2017} }
TY - EJOUR
T1 - SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION
AU - Wei Xu; Haoyuan Chi; Lei Zhou; Xiaolin Huang; Jie Yang
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1854
ER -
Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang. (2017). SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION. IEEE SigPort. http://sigport.org/1854
Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang, 2017. SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION. Available at: http://sigport.org/1854.
Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang. (2017). "SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION." Web.
1. Wei Xu, Haoyuan Chi, Lei Zhou, Xiaolin Huang, Jie Yang. SELF-PACED LEAST SQUARE SEMI-COUPLED DICTIONARY LEARNING FOR PERSON RE-IDENTIFICATION [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1854

Pages