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

Pattern recognition and classification (MLR-PATT)

A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions

Paper Details

Authors:
Submitted On:
14 October 2019 - 10:57am
Short Link:
Type:
Event:
Presenter's Name:
Document Year:
Cite

Document Files

Mouath_Aouayeb_ieee_mlsp2019.pdf

(3)

Subscribe

[1] , "A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4870. Accessed: Oct. 17, 2019.
@article{4870-19,
url = {http://sigport.org/4870},
author = { },
publisher = {IEEE SigPort},
title = {A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions},
year = {2019} }
TY - EJOUR
T1 - A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions
AU -
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4870
ER -
. (2019). A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions. IEEE SigPort. http://sigport.org/4870
, 2019. A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions. Available at: http://sigport.org/4870.
. (2019). "A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions." Web.
1. . A Spatiotemporal Deep Learning Solution For Automatic Micro-Expressions Recognition From Local Facial Regions [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4870

Robust importance-weighted cross-validation under sample selection bias


Cross-validation under sample selection bias can, in principle, be done by importance-weighting the empirical risk. However, the importance-weighted risk estimator produces sub-optimal hyperparameter estimates in problem settings where large weights arise with high probability. We study its sampling variance as a function of the training data distribution and introduce a control variate to increase its robustness to problematically large weights.

Paper Details

Authors:
Wouter M Kouw, Jesse H Krijthe, Marco Loog
Submitted On:
11 October 2019 - 4:48pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

PDF of poster

(5)

Subscribe

[1] Wouter M Kouw, Jesse H Krijthe, Marco Loog, "Robust importance-weighted cross-validation under sample selection bias", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4857. Accessed: Oct. 17, 2019.
@article{4857-19,
url = {http://sigport.org/4857},
author = {Wouter M Kouw; Jesse H Krijthe; Marco Loog },
publisher = {IEEE SigPort},
title = {Robust importance-weighted cross-validation under sample selection bias},
year = {2019} }
TY - EJOUR
T1 - Robust importance-weighted cross-validation under sample selection bias
AU - Wouter M Kouw; Jesse H Krijthe; Marco Loog
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4857
ER -
Wouter M Kouw, Jesse H Krijthe, Marco Loog. (2019). Robust importance-weighted cross-validation under sample selection bias. IEEE SigPort. http://sigport.org/4857
Wouter M Kouw, Jesse H Krijthe, Marco Loog, 2019. Robust importance-weighted cross-validation under sample selection bias. Available at: http://sigport.org/4857.
Wouter M Kouw, Jesse H Krijthe, Marco Loog. (2019). "Robust importance-weighted cross-validation under sample selection bias." Web.
1. Wouter M Kouw, Jesse H Krijthe, Marco Loog. Robust importance-weighted cross-validation under sample selection bias [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4857

A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense


While data poisoning attacks on classifiers were originally proposed to degrade a classifier's usability, there has been strong recent interest in backdoor data poisoning attacks, where the classifier learns to classify to a target class whenever a backdoor pattern ({\it e.g.}, a watermark or innocuous pattern) is added to an example from some class other than the target class.

Paper Details

Authors:
George Kesidis
Submitted On:
11 October 2019 - 1:35pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

MLSP19 backdoor poster 1.1.pdf

(8)

Subscribe

[1] George Kesidis, "A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4855. Accessed: Oct. 17, 2019.
@article{4855-19,
url = {http://sigport.org/4855},
author = {George Kesidis },
publisher = {IEEE SigPort},
title = {A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense},
year = {2019} }
TY - EJOUR
T1 - A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense
AU - George Kesidis
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4855
ER -
George Kesidis. (2019). A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense. IEEE SigPort. http://sigport.org/4855
George Kesidis, 2019. A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense. Available at: http://sigport.org/4855.
George Kesidis. (2019). "A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense." Web.
1. George Kesidis. A Benchmark Study of Backdoor Data Poisoning Defenses for Deep Neural Network Classifiers and A Novel Defense [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4855

3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion

Paper Details

Authors:
Submitted On:
27 September 2019 - 9:02am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

MMSP2019.pdf

(10)

Subscribe

[1] , "3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4846. Accessed: Oct. 17, 2019.
@article{4846-19,
url = {http://sigport.org/4846},
author = { },
publisher = {IEEE SigPort},
title = {3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion},
year = {2019} }
TY - EJOUR
T1 - 3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion
AU -
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4846
ER -
. (2019). 3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion. IEEE SigPort. http://sigport.org/4846
, 2019. 3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion. Available at: http://sigport.org/4846.
. (2019). "3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion." Web.
1. . 3D Facial Expression Recognition Based on Multi-View and Prior Knowledge Fusion [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4846

Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval


One of the key challenges of deep learning based image retrieval remains in aggregating convolutional activations into one highly representative feature vector. Ideally, this descriptor should encode semantic, spatial and low level information. Even though off-the-shelf pre-trained neural networks can already produce good representations in combination with aggregation methods, appropriate fine tuning for the task of image retrieval has shown to significantly boost retrieval performance.

Paper Details

Authors:
Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung
Submitted On:
25 September 2019 - 8:15am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

MMSP19_DARAC.pdf

(8)

Subscribe

[1] Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung, "Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4840. Accessed: Oct. 17, 2019.
@article{4840-19,
url = {http://sigport.org/4840},
author = {Konstantin Schall; Kai Uwe Barthel; Nico Hezel; Klaus Jung },
publisher = {IEEE SigPort},
title = {Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval},
year = {2019} }
TY - EJOUR
T1 - Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval
AU - Konstantin Schall; Kai Uwe Barthel; Nico Hezel; Klaus Jung
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4840
ER -
Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung. (2019). Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval. IEEE SigPort. http://sigport.org/4840
Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung, 2019. Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval. Available at: http://sigport.org/4840.
Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung. (2019). "Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval." Web.
1. Konstantin Schall, Kai Uwe Barthel, Nico Hezel, Klaus Jung. Deep Aggregation of Regional Convolutional Activations for Content Based Image Retrieval [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4840

DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING


Image-based people counting is a challenging work due to the large scale variation problem caused by the diversity of distance between the camera and the person, especially in the congested scenes. To handle this problem, the previous methods focus on building complicated models and rely on labeling the sophisticated density maps to learn the scale variation implicitly. It is often time-consuming in data pre-processing and difficult to train these deep models due to the lack of training data.

Paper Details

Authors:
Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang
Submitted On:
20 September 2019 - 10:06am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Crowd_Counting_ICIP_poster.pdf

(7)

Subscribe

[1] Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang, "DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4782. Accessed: Oct. 17, 2019.
@article{4782-19,
url = {http://sigport.org/4782},
author = {Van-Su Huynh; Vu-Hoang Tran; Ching-Chun Huang },
publisher = {IEEE SigPort},
title = {DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING},
year = {2019} }
TY - EJOUR
T1 - DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING
AU - Van-Su Huynh; Vu-Hoang Tran; Ching-Chun Huang
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4782
ER -
Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang. (2019). DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING. IEEE SigPort. http://sigport.org/4782
Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang, 2019. DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING. Available at: http://sigport.org/4782.
Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang. (2019). "DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING." Web.
1. Van-Su Huynh, Vu-Hoang Tran, Ching-Chun Huang. DANET: DEPTH-AWARE NETWORK FOR CROWD COUNTING [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4782

Photorealistic Image Synthesis for Object Instance Detection


We present an approach to synthesize highly photorealistic images of 3D object models, which we use to train a convolutional neural network for detecting the objects in real images. The proposed approach has three key ingredients: (1) 3D object models are rendered in 3D models of complete scenes with realistic materials and lighting, (2) plausible geometric configuration of objects and cameras in a scene is generated using physics simulation, and (3) high photorealism of the synthesized images is achieved by physically based rendering.

Paper Details

Authors:
Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter
Submitted On:
19 September 2019 - 7:59am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

objectsynth_icip2019_slides.pdf

(5)

Subscribe

[1] Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter, "Photorealistic Image Synthesis for Object Instance Detection", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4726. Accessed: Oct. 17, 2019.
@article{4726-19,
url = {http://sigport.org/4726},
author = {Tomas Hodan; Vibhav Vineet; Ran Gal; Emanuel Shalev; Jon Hanzelka; Treb Connell; Pedro Urbina; Sudipta N. Sinha; Brian Guenter },
publisher = {IEEE SigPort},
title = {Photorealistic Image Synthesis for Object Instance Detection},
year = {2019} }
TY - EJOUR
T1 - Photorealistic Image Synthesis for Object Instance Detection
AU - Tomas Hodan; Vibhav Vineet; Ran Gal; Emanuel Shalev; Jon Hanzelka; Treb Connell; Pedro Urbina; Sudipta N. Sinha; Brian Guenter
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4726
ER -
Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter. (2019). Photorealistic Image Synthesis for Object Instance Detection. IEEE SigPort. http://sigport.org/4726
Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter, 2019. Photorealistic Image Synthesis for Object Instance Detection. Available at: http://sigport.org/4726.
Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter. (2019). "Photorealistic Image Synthesis for Object Instance Detection." Web.
1. Tomas Hodan, Vibhav Vineet, Ran Gal, Emanuel Shalev, Jon Hanzelka, Treb Connell, Pedro Urbina, Sudipta N. Sinha, Brian Guenter. Photorealistic Image Synthesis for Object Instance Detection [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4726

MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION


With the development of deep learning, many state-of-the-art natural image scene classification methods have demonstrated impressive performance. While the current convolution neural network tends to extract global features and global semantic information in a scene, the geo-spatial objects can be located at anywhere in an aerial image scene and their spatial arrangement tends to be more complicated. One possible solution is to preserve more local semantic information and enhance feature propagation.

Paper Details

Authors:
Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu
Submitted On:
19 September 2019 - 1:10am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICIP1792_Poster.pdf

(6)

Subscribe

[1] Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu, "MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4707. Accessed: Oct. 17, 2019.
@article{4707-19,
url = {http://sigport.org/4707},
author = {Qi Bi; Kun Qin; Zhili Li; Han Zhang; Kai Xu },
publisher = {IEEE SigPort},
title = {MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION},
year = {2019} }
TY - EJOUR
T1 - MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION
AU - Qi Bi; Kun Qin; Zhili Li; Han Zhang; Kai Xu
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4707
ER -
Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu. (2019). MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION. IEEE SigPort. http://sigport.org/4707
Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu, 2019. MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION. Available at: http://sigport.org/4707.
Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu. (2019). "MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION." Web.
1. Qi Bi, Kun Qin, Zhili Li, Han Zhang, Kai Xu. MULTIPLE INSTANCE DENSE CONNECTED CONVOLUTION NEURAL NETWORK FOR AERIAL IMAGE SCENE CLASSIFICATION [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4707

Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications


Performance of 6DoF pose estimation techniques from RGB/RGB-D images has improved significantly with sophisticated deep learning frameworks. These frameworks require large-scale training data based on real/synthetic RGB/RGB-D information. Difficulty of obtaining adequate training data has limited the scope of these frameworks for ubiquitous application areas. Also, fast pose estimation at inference time often requires high-end GPU(s) that restricts the scope for its application in mobile hardware.

Paper Details

Authors:
Bowen Chen, Juhan Bae, Dibyendu Mukherjee
Submitted On:
18 September 2019 - 1:28pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

FAST 6DOF POSE ESTIMATION WITH SYNTHETIC TEXTURELESS CAD MODEL FOR MOBILE APPLICATIONS - Poster ICIP_Static.pdf

(61)

Subscribe

[1] Bowen Chen, Juhan Bae, Dibyendu Mukherjee, "Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4690. Accessed: Oct. 17, 2019.
@article{4690-19,
url = {http://sigport.org/4690},
author = {Bowen Chen; Juhan Bae; Dibyendu Mukherjee },
publisher = {IEEE SigPort},
title = {Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications},
year = {2019} }
TY - EJOUR
T1 - Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications
AU - Bowen Chen; Juhan Bae; Dibyendu Mukherjee
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4690
ER -
Bowen Chen, Juhan Bae, Dibyendu Mukherjee. (2019). Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications. IEEE SigPort. http://sigport.org/4690
Bowen Chen, Juhan Bae, Dibyendu Mukherjee, 2019. Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications. Available at: http://sigport.org/4690.
Bowen Chen, Juhan Bae, Dibyendu Mukherjee. (2019). "Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications." Web.
1. Bowen Chen, Juhan Bae, Dibyendu Mukherjee. Fast 6dof Pose Estimation with Synthetic Textureless Cad Model for Mobile Applications [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4690

Detecting Arbitrarily Rotated Faces for Face Analysis


Current face detection concentrates on detecting tiny faces and severely occluded faces. Face analysis methods, however, require a good localization and would benefit greatly from some rotation information. We propose to predict a face direction vector (FDV), which provides the face size and orientation and can be learned by a common object detection architecture better than the traditional bounding box. It provides a more consistent definition of face location and size. Using the FDV is promising for all succeeding face analysis methods.

Paper Details

Authors:
Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi
Submitted On:
18 September 2019 - 12:38pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Saxen19 - Detecting Arbitrarily Rotated Faces - Poster.pptx

(10)

Keywords

Additional Categories

Subscribe

[1] Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi, "Detecting Arbitrarily Rotated Faces for Face Analysis", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4689. Accessed: Oct. 17, 2019.
@article{4689-19,
url = {http://sigport.org/4689},
author = {Frerk Saxen; Sebastian Handrich; Philipp Werner; Ehsan Othman; Ayoub Al-Hamadi },
publisher = {IEEE SigPort},
title = {Detecting Arbitrarily Rotated Faces for Face Analysis},
year = {2019} }
TY - EJOUR
T1 - Detecting Arbitrarily Rotated Faces for Face Analysis
AU - Frerk Saxen; Sebastian Handrich; Philipp Werner; Ehsan Othman; Ayoub Al-Hamadi
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4689
ER -
Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi. (2019). Detecting Arbitrarily Rotated Faces for Face Analysis. IEEE SigPort. http://sigport.org/4689
Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi, 2019. Detecting Arbitrarily Rotated Faces for Face Analysis. Available at: http://sigport.org/4689.
Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi. (2019). "Detecting Arbitrarily Rotated Faces for Face Analysis." Web.
1. Frerk Saxen, Sebastian Handrich, Philipp Werner, Ehsan Othman, Ayoub Al-Hamadi. Detecting Arbitrarily Rotated Faces for Face Analysis [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4689

Pages