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

Machine Learning for Signal Processing

Sparse Modeling


Sparse Modeling in Image Processing and Deep LearningSparse approximation is a well-established theory, with a profound impact on the fields of signal and image processing. In this talk we start by presenting this model and its features, and then turn to describe two special cases of it – the convolutional sparse coding (CSC) and its multi-layered version (ML-CSC).  Amazingly, as we will carefully show, ML-CSC provides a solid theoretical foundation to … deep-learning.

Paper Details

Authors:
Michael Elad
Submitted On:
22 December 2017 - 1:26pm
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

ICIP_KeyNote_Talk_small size.pdf

(49)

Subscribe

[1] Michael Elad, "Sparse Modeling ", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/2260. Accessed: Sep. 20, 2019.
@article{2260-17,
url = {http://sigport.org/2260},
author = {Michael Elad },
publisher = {IEEE SigPort},
title = {Sparse Modeling },
year = {2017} }
TY - EJOUR
T1 - Sparse Modeling
AU - Michael Elad
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/2260
ER -
Michael Elad. (2017). Sparse Modeling . IEEE SigPort. http://sigport.org/2260
Michael Elad, 2017. Sparse Modeling . Available at: http://sigport.org/2260.
Michael Elad. (2017). "Sparse Modeling ." Web.
1. Michael Elad. Sparse Modeling [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/2260

A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING


Convolutional neural networks (CNN) have shown state-of-the-art results for low-level computer vision problems such as stereo and monocular disparity estimations, but still, have much room to further improve their performance in terms of accuracy, numbers of parameters, etc. Recent works have uncovered the advantages of using an unsupervised scheme to train CNN’s to estimate monocular disparity, where only the relatively-easy-to-obtain stereo images are needed for training.

Paper Details

Authors:
Munchurl Kim
Submitted On:
19 September 2019 - 8:16am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:

Document Files

Poster_1748.pdf

(1)

Subscribe

[1] Munchurl Kim, "A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4712. Accessed: Sep. 20, 2019.
@article{4712-19,
url = {http://sigport.org/4712},
author = {Munchurl Kim },
publisher = {IEEE SigPort},
title = {A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING},
year = {2019} }
TY - EJOUR
T1 - A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING
AU - Munchurl Kim
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4712
ER -
Munchurl Kim. (2019). A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING. IEEE SigPort. http://sigport.org/4712
Munchurl Kim, 2019. A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING. Available at: http://sigport.org/4712.
Munchurl Kim. (2019). "A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING." Web.
1. Munchurl Kim. A NOVEL MONOCULAR DISPARITY ESTIMATION NETWORK WITH DOMAIN TRANSFORMATION AND AMBIGUITY LEARNING [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4712

MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS

Paper Details

Authors:
Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee
Submitted On:
19 September 2019 - 2:46am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Elkhamy_Telewide_depth_2019_ICIP.pdf

(3)

Subscribe

[1] Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee, "MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4709. Accessed: Sep. 20, 2019.
@article{4709-19,
url = {http://sigport.org/4709},
author = {Mostafa El-Khamy; Xianzhi Du; Haoyu Ren; Jungwon Lee },
publisher = {IEEE SigPort},
title = {MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS},
year = {2019} }
TY - EJOUR
T1 - MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS
AU - Mostafa El-Khamy; Xianzhi Du; Haoyu Ren; Jungwon Lee
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4709
ER -
Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee. (2019). MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS. IEEE SigPort. http://sigport.org/4709
Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee, 2019. MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS. Available at: http://sigport.org/4709.
Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee. (2019). "MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS." Web.
1. Mostafa El-Khamy, Xianzhi Du, Haoyu Ren, Jungwon Lee. MULTI TASK LEARNING OF DEPTH FROM TELE AND WIDE STEREO IMAGE PAIRS [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4709

PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING


Portrait segmentation is becoming a hot topic nowadays.
In this paper we propose a novel framework to cope with
the high precision requirements that portrait segmentation
demands on boundary area by deep refinement of the
portrait matting. Our approach introduces three novel
techniques. First, a trimap is proposed by fusing information
coming from two well-known techniques for image
segmentation, i.e., Mask R-CNN and DensePose. Second,
an alpha matting algorithm runs over the previous trimap

Paper Details

Authors:
Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán
Submitted On:
17 September 2019 - 6:00am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

poster paper code 2790

(6)

Keywords

Additional Categories

Subscribe

[1] Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán, "PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4658. Accessed: Sep. 20, 2019.
@article{4658-19,
url = {http://sigport.org/4658},
author = {Carlos Orrite; Miguel Angel Varona; Eduardo Estopiñán; José Ramón Beltrán },
publisher = {IEEE SigPort},
title = {PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING},
year = {2019} }
TY - EJOUR
T1 - PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING
AU - Carlos Orrite; Miguel Angel Varona; Eduardo Estopiñán; José Ramón Beltrán
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4658
ER -
Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán. (2019). PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING. IEEE SigPort. http://sigport.org/4658
Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán, 2019. PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING. Available at: http://sigport.org/4658.
Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán. (2019). "PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING." Web.
1. Carlos Orrite, Miguel Angel Varona, Eduardo Estopiñán, José Ramón Beltrán. PORTRAIT SEGMENTATION BY DEEP REFINEMENT OF IMAGE MATTING [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4658

DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS


Even though zero padding is usually a staple in convolutional
neural networks to maintain the output size, it is highly suspicious
because it significantly alters the input distribution
around border region. To mitigate this problem, in this paper,
we propose a new padding technique termed as distribution
padding. The goal of the method is to approximately maintain
the statistics of the input border regions. We introduce
two different ways to achieve our goal. In both approaches,
the padded values are derived from the means of the border

Paper Details

Authors:
Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee
Submitted On:
17 September 2019 - 3:06am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

distributional padding.pptx

(4)

Subscribe

[1] Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee, "DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS ", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4654. Accessed: Sep. 20, 2019.
@article{4654-19,
url = {http://sigport.org/4654},
author = {Anh-Duc Nguyen; Seonghwa Choi; Woojae Kim; Sewoong Ahn; Jinwoo Kim; Sanghoon Lee },
publisher = {IEEE SigPort},
title = {DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS },
year = {2019} }
TY - EJOUR
T1 - DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS
AU - Anh-Duc Nguyen; Seonghwa Choi; Woojae Kim; Sewoong Ahn; Jinwoo Kim; Sanghoon Lee
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4654
ER -
Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee. (2019). DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS . IEEE SigPort. http://sigport.org/4654
Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee, 2019. DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS . Available at: http://sigport.org/4654.
Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee. (2019). "DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS ." Web.
1. Anh-Duc Nguyen, Seonghwa Choi, Woojae Kim, Sewoong Ahn, Jinwoo Kim, Sanghoon Lee. DISTRIBUTION PADDING IN CONVOLUTIONAL NEURAL NETWORKS [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4654

A History-based Stopping Criterion in Recursive Bayesian State Estimation


In dynamic state-space models, the state can be estimated through recursive computation of the posterior distribution of the state given all measurements. In scenarios where active sensing/querying is possible, a hard decision is made when the state posterior achieves a pre-set confidence threshold. This mandate to meet a hard threshold may sometimes unnecessarily require more queries. In application domains where sensing/querying cost is of concern, some potential accuracy may be sacrificed for greater gains in sensing cost.

Paper Details

Authors:
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus
Submitted On:
15 May 2019 - 9:57pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

P_Marghi_Kocan_ICASSP_2019.pdf

(34)

Subscribe

[1] Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus, "A History-based Stopping Criterion in Recursive Bayesian State Estimation", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4536. Accessed: Sep. 20, 2019.
@article{4536-19,
url = {http://sigport.org/4536},
author = {Yeganeh M. Marghi; Aziz Kocanaogullari; Murat Akcakaya; Deniz Erdomus },
publisher = {IEEE SigPort},
title = {A History-based Stopping Criterion in Recursive Bayesian State Estimation},
year = {2019} }
TY - EJOUR
T1 - A History-based Stopping Criterion in Recursive Bayesian State Estimation
AU - Yeganeh M. Marghi; Aziz Kocanaogullari; Murat Akcakaya; Deniz Erdomus
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4536
ER -
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. (2019). A History-based Stopping Criterion in Recursive Bayesian State Estimation. IEEE SigPort. http://sigport.org/4536
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus, 2019. A History-based Stopping Criterion in Recursive Bayesian State Estimation. Available at: http://sigport.org/4536.
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. (2019). "A History-based Stopping Criterion in Recursive Bayesian State Estimation." Web.
1. Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. A History-based Stopping Criterion in Recursive Bayesian State Estimation [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4536

A History-based Stopping Criterion in Recursive Bayesian State Estimation


In dynamic state-space models, the state can be estimated through recursive computation of the posterior distribution of the state given all measurements. In scenarios where active sensing/querying is possible, a hard decision is made when the state posterior achieves a pre-set confidence threshold. This mandate to meet a hard threshold may sometimes unnecessarily require more queries. In application domains where sensing/querying cost is of concern, some potential accuracy may be sacrificed for greater gains in sensing cost.

Paper Details

Authors:
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus
Submitted On:
15 May 2019 - 9:57pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

P_Marghi_Kocan_ICASSP_2019.pdf

(32)

Subscribe

[1] Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus, "A History-based Stopping Criterion in Recursive Bayesian State Estimation", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4535. Accessed: Sep. 20, 2019.
@article{4535-19,
url = {http://sigport.org/4535},
author = {Yeganeh M. Marghi; Aziz Kocanaogullari; Murat Akcakaya; Deniz Erdomus },
publisher = {IEEE SigPort},
title = {A History-based Stopping Criterion in Recursive Bayesian State Estimation},
year = {2019} }
TY - EJOUR
T1 - A History-based Stopping Criterion in Recursive Bayesian State Estimation
AU - Yeganeh M. Marghi; Aziz Kocanaogullari; Murat Akcakaya; Deniz Erdomus
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4535
ER -
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. (2019). A History-based Stopping Criterion in Recursive Bayesian State Estimation. IEEE SigPort. http://sigport.org/4535
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus, 2019. A History-based Stopping Criterion in Recursive Bayesian State Estimation. Available at: http://sigport.org/4535.
Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. (2019). "A History-based Stopping Criterion in Recursive Bayesian State Estimation." Web.
1. Yeganeh M. Marghi, Aziz Kocanaogullari, Murat Akcakaya, Deniz Erdomus. A History-based Stopping Criterion in Recursive Bayesian State Estimation [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4535

Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks

Paper Details

Authors:
Sumit Jha, Carlos Busso
Submitted On:
13 May 2019 - 9:22am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

icassp-poster.pdf

(32)

Subscribe

[1] Sumit Jha, Carlos Busso, "Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4485. Accessed: Sep. 20, 2019.
@article{4485-19,
url = {http://sigport.org/4485},
author = {Sumit Jha; Carlos Busso },
publisher = {IEEE SigPort},
title = {Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks},
year = {2019} }
TY - EJOUR
T1 - Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks
AU - Sumit Jha; Carlos Busso
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4485
ER -
Sumit Jha, Carlos Busso. (2019). Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks. IEEE SigPort. http://sigport.org/4485
Sumit Jha, Carlos Busso, 2019. Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks. Available at: http://sigport.org/4485.
Sumit Jha, Carlos Busso. (2019). "Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks." Web.
1. Sumit Jha, Carlos Busso. Estimation of Gaze Region using Two Dimensional Probabilistic Maps Constructed using Convolutional Neural Networks [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4485

Adversarial Speaker Adaptation


We propose a novel adversarial speaker adaptation (ASA) scheme, in which adversarial learning is applied to regularize the distribution of deep hidden features in a speaker-dependent (SD) deep neural network (DNN) acoustic model to be close to that of a fixed speaker-independent (SI) DNN acoustic model during adaptation. An additional discriminator network is introduced to distinguish the deep features generated by the SD model from those produced by the SI model.

Paper Details

Authors:
Zhong Meng, Jinyu Li, Yifan Gong
Submitted On:
12 May 2019 - 9:26pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

asa_oral_v3.pptx

(48)

Subscribe

[1] Zhong Meng, Jinyu Li, Yifan Gong, "Adversarial Speaker Adaptation", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4475. Accessed: Sep. 20, 2019.
@article{4475-19,
url = {http://sigport.org/4475},
author = {Zhong Meng; Jinyu Li; Yifan Gong },
publisher = {IEEE SigPort},
title = {Adversarial Speaker Adaptation},
year = {2019} }
TY - EJOUR
T1 - Adversarial Speaker Adaptation
AU - Zhong Meng; Jinyu Li; Yifan Gong
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4475
ER -
Zhong Meng, Jinyu Li, Yifan Gong. (2019). Adversarial Speaker Adaptation. IEEE SigPort. http://sigport.org/4475
Zhong Meng, Jinyu Li, Yifan Gong, 2019. Adversarial Speaker Adaptation. Available at: http://sigport.org/4475.
Zhong Meng, Jinyu Li, Yifan Gong. (2019). "Adversarial Speaker Adaptation." Web.
1. Zhong Meng, Jinyu Li, Yifan Gong. Adversarial Speaker Adaptation [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4475

Attentive Adversarial Learning for Domain-Invariant Training


Adversarial domain-invariant training (ADIT) proves to be effective in suppressing the effects of domain variability in acoustic modeling and has led to improved performance in automatic speech recognition (ASR). In ADIT, an auxiliary domain classifier takes in equally-weighted deep features from a deep neural network (DNN) acoustic model and is trained to improve their domain-invariance by optimizing an adversarial loss function.

Paper Details

Authors:
Zhong Meng, Jinyu Li, Yifan Gong
Submitted On:
12 May 2019 - 9:03pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

aadit_poster.pptx

(33)

Subscribe

[1] Zhong Meng, Jinyu Li, Yifan Gong, "Attentive Adversarial Learning for Domain-Invariant Training", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4474. Accessed: Sep. 20, 2019.
@article{4474-19,
url = {http://sigport.org/4474},
author = {Zhong Meng; Jinyu Li; Yifan Gong },
publisher = {IEEE SigPort},
title = {Attentive Adversarial Learning for Domain-Invariant Training},
year = {2019} }
TY - EJOUR
T1 - Attentive Adversarial Learning for Domain-Invariant Training
AU - Zhong Meng; Jinyu Li; Yifan Gong
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4474
ER -
Zhong Meng, Jinyu Li, Yifan Gong. (2019). Attentive Adversarial Learning for Domain-Invariant Training. IEEE SigPort. http://sigport.org/4474
Zhong Meng, Jinyu Li, Yifan Gong, 2019. Attentive Adversarial Learning for Domain-Invariant Training. Available at: http://sigport.org/4474.
Zhong Meng, Jinyu Li, Yifan Gong. (2019). "Attentive Adversarial Learning for Domain-Invariant Training." Web.
1. Zhong Meng, Jinyu Li, Yifan Gong. Attentive Adversarial Learning for Domain-Invariant Training [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4474

Pages