Bong-Nam Kang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Bong-Nam Kang has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
Daijin Kim
15 shared publications- Daijin Kim
Pohang University of Science and Technology (POSTECH)
7 shared publications Yonghyun Kim
6 shared publications- Yonghyun Kim
Kakao Enterprise
4 shared publications Hai Wang
3 shared publicationsJongmin Yoon
Pohang University of Science and Technology (POSTECH), Pohang, Republic of Korea
3 shared publications
Latest Publications
- 1 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 5471-5480
Attentional Feature-Pair Relation Networks for Accurate Face Recognition
- 1 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 5471-5480
Attentional Feature-Pair Relation Networks for Accurate Face Recognition
- 26 May 2019
- Published by Springer Nature
- p. 555-570
BAN: Focusing on Boundary Context for Object Detection
- 26 May 2019
- Published by Springer Nature
- p. 555-570
BAN: Focusing on Boundary Context for Object Detection
- Open Access
Pairwise Relational Networks using Local Appearance Features for Face Recognition
We propose a new face recognition method, called a pairwise relational network (PRN), which takes local appearance features around landmark points on the feature map, and captures unique pairwise relations with the same identity and discriminative pairwise relations between different identities. The PRN aims to determine facial part-relational structure from local appearance feature pairs. Because meaningful pairwise relations should be identity dependent, we add a face identity state feature, which obtains from the long short-term memory (LSTM) units network with the sequential local appearance features. To further improve accuracy, we combined the global appearance features with the pairwise relational feature. Experimental results on the LFW show that the PRN achieved 99.76% accuracy. On the YTF, PRN achieved the state-of-the-art accuracy (96.3%). The PRN also achieved comparable results to the state-of-the-art for both face verification and face identification tasks on the IJB-A and IJB-B. This work is already published on ECCV 2018.
Open AccessPairwise Relational Networks using Local Appearance Features for Face Recognition
We propose a new face recognition method, called a pairwise relational network (PRN), which takes local appearance features around landmark points on the feature map, and captures unique pairwise relations with the same identity and discriminative pairwise relations between different identities. The PRN aims to determine facial part-relational structure from local appearance feature pairs. Because meaningful pairwise relations should be identity dependent, we add a face identity state feature, which obtains from the long short-term memory (LSTM) units network with the sequential local appearance features. To further improve accuracy, we combined the global appearance features with the pairwise relational feature. Experimental results on the LFW show that the PRN achieved 99.76% accuracy. On the YTF, PRN achieved the state-of-the-art accuracy (96.3%). The PRN also achieved comparable results to the state-of-the-art for both face verification and face identification tasks on the IJB-A and IJB-B. This work is already published on ECCV 2018.
- Open Access
BAN: Focusing on Boundary Context for Object Detection
Visual context is one of the important clue for object detection and the context information for boundaries of an object is especially valuable. We propose a boundary aware network (BAN) designed to exploit the visual contexts including boundary information and surroundings, named boundary context, and define three types of the boundary contexts: side, vertex and in/out-boundary context. Our BAN consists of 10 sub-networks for the area belonging to the boundary contexts. The detection head of BAN is defined as an ensemble of these sub-networks with different contributions depending on the sub-problem of detection. To verify our method, we visualize the activation of the sub-networks according to the boundary contexts and empirically show that the sub-networks contribute more to the related sub-problem in detection. We evaluate our method on PASCAL VOC detection benchmark and MS COCO dataset. The proposed method achieves the mean Average Precision (mAP) of 83.4% on PASCAL VOC and 36.9% on MS COCO. BAN allows the convolution network to provide an additional source of contexts for detection and selectively focus on the more important contexts, and it can be generally applied to many other detection methods as well to enhance the accuracy in detection.
Open AccessBAN: Focusing on Boundary Context for Object Detection
Visual context is one of the important clue for object detection and the context information for boundaries of an object is especially valuable. We propose a boundary aware network (BAN) designed to exploit the visual contexts including boundary information and surroundings, named boundary context, and define three types of the boundary contexts: side, vertex and in/out-boundary context. Our BAN consists of 10 sub-networks for the area belonging to the boundary contexts. The detection head of BAN is defined as an ensemble of these sub-networks with different contributions depending on the sub-problem of detection. To verify our method, we visualize the activation of the sub-networks according to the boundary contexts and empirically show that the sub-networks contribute more to the related sub-problem in detection. We evaluate our method on PASCAL VOC detection benchmark and MS COCO dataset. The proposed method achieves the mean Average Precision (mAP) of 83.4% on PASCAL VOC and 36.9% on MS COCO. BAN allows the convolution network to provide an additional source of contexts for detection and selectively focus on the more important contexts, and it can be generally applied to many other detection methods as well to enhance the accuracy in detection.
- 9 October 2018
- Published by Springer Nature
- p. 646-663
Pairwise Relational Networks for Face Recognition
- 9 October 2018
- Published by Springer Nature
- p. 646-663
Pairwise Relational Networks for Face Recognition
- 6 October 2018
- Published by Springer Nature
- p. 328-343
SAN: Learning Relationship Between Convolutional Features for Multi-scale Object Detection
- 6 October 2018
- Published by Springer Nature
- p. 328-343
SAN: Learning Relationship Between Convolutional Features for Multi-scale Object Detection
- 1 September 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 420-424
Detector with focus: Normalizing gradient in image pyramid
An image pyramid can extend many object detection algorithms to solve detection on multiple scales. However, interpolation during the resampling process of an image pyramid causes gradient variation, which is the difference of the gradients between the original image and the scaled images. Our key insight is that the increased variance of gradients makes the classifiers have difficulty in correctly assigning categories. We prove the existence of the gradient variation by formulating the ratio of gradient expectations between an original image and scaled images, then propose a simple and novel gradient normalization method to eliminate the effect of this variation. The proposed normalization method reduce the variance in an image pyramid and allow the classifier to focus on a smaller coverage. We show the improvement in three different visual recognition problems: pedestrian detection, pose estimation, and object detection. The method is generally applicable to many vision algorithms based on an image pyramid with gradients.
- 1 September 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 420-424
Detector with focus: Normalizing gradient in image pyramid
An image pyramid can extend many object detection algorithms to solve detection on multiple scales. However, interpolation during the resampling process of an image pyramid causes gradient variation, which is the difference of the gradients between the original image and the scaled images. Our key insight is that the increased variance of gradients makes the classifiers have difficulty in correctly assigning categories. We prove the existence of the gradient variation by formulating the ratio of gradient expectations between an original image and scaled images, then propose a simple and novel gradient normalization method to eliminate the effect of this variation. The proposed normalization method reduce the variance in an image pyramid and allow the classifier to focus on a smaller coverage. We show the improvement in three different visual recognition problems: pedestrian detection, pose estimation, and object detection. The method is generally applicable to many vision algorithms based on an image pyramid with gradients.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 611-618
Deep Convolutional Neural Network Using Triplets of Faces, Deep Ensemble, and Score-Level Fusion for Face Recognition
This paper proposes a new face verification method that uses multiple deep convolutional neural networks (DCNNs) and a deep ensemble, that extracts two types of low dimensional but discriminative and high-level abstracted features from each DCNN, then combines them as a descriptor for face verification. Our DCNNs are built from stacked multi-scale convolutional layer blocks to present multi-scale abstraction. To train our DCNNs, we use different resolutions of triplets that consist of reference images, positive images, and negative images, and triplet-based loss function that maximize the ratio of distances between negative pairs and positive pairs and minimize the absolute distances between positive face images. A deep ensemble is generated from features extracted by each DCNN, and used as a descriptor to train the joint Bayesian learning and its transfer learning method. On the LFW, although we use only 198,018 images and only four different types of networks, the proposed method with the joint Bayesian learning and its transfer learning method achieved 98.33% accuracy. In addition to further increase the accuracy, we combine the proposed method and high dimensional LBP based joint Bayesian method, and achieved 99.08% accuracy on the LFW. Therefore, the proposed method helps to improve the accuracy of face verification when training data is insufficient to train DCNNs.
- 1 July 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 611-618
Deep Convolutional Neural Network Using Triplets of Faces, Deep Ensemble, and Score-Level Fusion for Face Recognition
This paper proposes a new face verification method that uses multiple deep convolutional neural networks (DCNNs) and a deep ensemble, that extracts two types of low dimensional but discriminative and high-level abstracted features from each DCNN, then combines them as a descriptor for face verification. Our DCNNs are built from stacked multi-scale convolutional layer blocks to present multi-scale abstraction. To train our DCNNs, we use different resolutions of triplets that consist of reference images, positive images, and negative images, and triplet-based loss function that maximize the ratio of distances between negative pairs and positive pairs and minimize the absolute distances between positive face images. A deep ensemble is generated from features extracted by each DCNN, and used as a descriptor to train the joint Bayesian learning and its transfer learning method. On the LFW, although we use only 198,018 images and only four different types of networks, the proposed method with the joint Bayesian learning and its transfer learning method achieved 98.33% accuracy. In addition to further increase the accuracy, we combine the proposed method and high dimensional LBP based joint Bayesian method, and achieved 99.08% accuracy on the LFW. Therefore, the proposed method helps to improve the accuracy of face verification when training data is insufficient to train DCNNs.
- 1 June 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 97-101
Reliable multi-person identification using DCNN-based face recognition algorithm and scale-ratio method
Recently, deep convolutional neural networks (DCNNs) have set a new trend in the computer vision community by improving the state-of-the-art performance in almost all of applications. We propose DCNN-based face recognition algorithm. This paper aims at analyzing and verifying considerations when the proposed method is implemented in a real environment. First, Multiple images of the same scene are processed. Also, this novel method is considered together with a method to reduce the total recognition time in the perspective of integrated system. We analyzed the experimental data and evaluated the performance by setting up the sensor fusion network in actual classroom.
- 1 June 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 97-101
Reliable multi-person identification using DCNN-based face recognition algorithm and scale-ratio method
Recently, deep convolutional neural networks (DCNNs) have set a new trend in the computer vision community by improving the state-of-the-art performance in almost all of applications. We propose DCNN-based face recognition algorithm. This paper aims at analyzing and verifying considerations when the proposed method is implemented in a real environment. First, Multiple images of the same scene are processed. Also, this novel method is considered together with a method to reduce the total recognition time in the perspective of integrated system. We analyzed the experimental data and evaluated the performance by setting up the sensor fusion network in actual classroom.
- 9 February 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 004460-004465
Deep convolution neural network with stacks of multi-scale convolutional layer block using triplet of faces for face recognition in the wild
Recently, deep convolutional neural networks have set a new trend in fields of face recognition by improving the state-of-the-art performance. By using deep neural networks, much more sophisticated and high level abstracted features can be learned automatically. In this paper, we propose a method for face recognition using multi-scale convolution layer blocks and triplets of faces in unconstrained environments. We use the ensemble of deep convolution neural networks trained on differently scaled and aligned face images. This extracts low dimensional but high-level abstraction and discriminative features for face recognition. With these features, we employ the jointly Bayesian model and transfer learning which adapts the knowledge trained from the source domain to target domain. Experiment shows that our proposed method achieves 98.33% pair-wise verification accuracy on the LFW dataset.
- 9 February 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 004460-004465
Deep convolution neural network with stacks of multi-scale convolutional layer block using triplet of faces for face recognition in the wild
Recently, deep convolutional neural networks have set a new trend in fields of face recognition by improving the state-of-the-art performance. By using deep neural networks, much more sophisticated and high level abstracted features can be learned automatically. In this paper, we propose a method for face recognition using multi-scale convolution layer blocks and triplets of faces in unconstrained environments. We use the ensemble of deep convolution neural networks trained on differently scaled and aligned face images. This extracts low dimensional but high-level abstraction and discriminative features for face recognition. With these features, we employ the jointly Bayesian model and transfer learning which adapts the knowledge trained from the source domain to target domain. Experiment shows that our proposed method achieves 98.33% pair-wise verification accuracy on the LFW dataset.