Network representation learning (NRL) is a crucial method to learn low-dimensional vertex representations to capture network information. However, conventional NRL models only regard each edge as a binary or continuous value while neglecting the rich semantic information on edges. To enhance network representation for Social Relation Extraction (SRE) task, we present a novel deep neural network based model, EsiNet, by learning the structure and semantic information of edges simultaneously. Compared with previous work, EsiNet focuses on further learning the interactions between vertices and capturing the correlations between labels. By jointly optimizing the objective function of these two components, EsiNet can preserve both the semantic and structural information of edges. Extensive experiments on several public datasets demonstrate that EsiNet outperforms other baselines significantly, by around 3% to 5% on hits@10 absolutely.
{"title":"EsiNet: Enhanced Network Representation via Further Learning the Semantic Information of Edges","authors":"Anqing Zheng, Chong Feng, Fang Yang, Huanhuan Zhang","doi":"10.1109/ICTAI.2019.00142","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00142","url":null,"abstract":"Network representation learning (NRL) is a crucial method to learn low-dimensional vertex representations to capture network information. However, conventional NRL models only regard each edge as a binary or continuous value while neglecting the rich semantic information on edges. To enhance network representation for Social Relation Extraction (SRE) task, we present a novel deep neural network based model, EsiNet, by learning the structure and semantic information of edges simultaneously. Compared with previous work, EsiNet focuses on further learning the interactions between vertices and capturing the correlations between labels. By jointly optimizing the objective function of these two components, EsiNet can preserve both the semantic and structural information of edges. Extensive experiments on several public datasets demonstrate that EsiNet outperforms other baselines significantly, by around 3% to 5% on hits@10 absolutely.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116551448","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00031
B. Jaumard, Kia Babashahi Ashtiani, Nicolas Huin
In the context of multi-agent systems, Automated Mechanism Design (AMD) is the computer-based design of the rules of a mechanism, which reaches an equilibrium despite the fact that agents can be selfish and lie about their preferences. Although it has been shown that AMD can be modelled as a linear program, it is with an exponential number of variables and consequently, there is no known efficient algorithm. We revisit the latter linear program model proposed for the AMD problem and introduce a new one with a polynomial number of variables. We show that the latter model corresponds to a Dantzig-Wolfe decomposition of the second one and design efficient solution schemes in polynomial time for both two models. Numerical experiments compare the solution efficiency of both models and show that we can solve very significantly larger data instances than before, up to 2,000 agents or 2,000 resources in about 35 seconds.
{"title":"Automated Mechanism Design: Compact and Decomposition Linear Programming Models","authors":"B. Jaumard, Kia Babashahi Ashtiani, Nicolas Huin","doi":"10.1109/ICTAI.2019.00031","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00031","url":null,"abstract":"In the context of multi-agent systems, Automated Mechanism Design (AMD) is the computer-based design of the rules of a mechanism, which reaches an equilibrium despite the fact that agents can be selfish and lie about their preferences. Although it has been shown that AMD can be modelled as a linear program, it is with an exponential number of variables and consequently, there is no known efficient algorithm. We revisit the latter linear program model proposed for the AMD problem and introduce a new one with a polynomial number of variables. We show that the latter model corresponds to a Dantzig-Wolfe decomposition of the second one and design efficient solution schemes in polynomial time for both two models. Numerical experiments compare the solution efficiency of both models and show that we can solve very significantly larger data instances than before, up to 2,000 agents or 2,000 resources in about 35 seconds.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114854129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00216
Nianwen Ning, Chenguang Song, Pengpeng Zhou, Yunlei Zhang, Bin Wu
Network embedding aims to learn a latent representation of each node which preserves the structure information. Many real-world networks have multiple dimensions of nodes and multiple types of relations. Therefore, it is more appropriate to represent such kind of networks as multiplex networks. A multiplex network is formed by a set of nodes connected in different layers by links indicating interactions of different types. However, existing random walk based multiplex networks embedding algorithms have problems with sampling bias and imbalanced relation types, thus leading the poor performance in the downstream tasks. In this paper, we propose a node embedding method based on adaptive cross-layer forest fire sampling (FFS) for multiplex networks (FFME). We first focus on the sampling strategies of FFS to address the bias issue of random walk. We utilize a fixed-length queue to record previously visited layers, which can balance the edge distribution over different layers in sampled node sequences. In addition, to adaptively sample node's context, we also propose a metric for node called Neighbors Partition Coefficient (N P C ). The generation process of node sequence is supervised by NPC for adaptive cross-layer sampling. Experiments on real-world networks in diverse fields show that our method outperforms the state-of-the-art methods in application tasks such as cross-domain link prediction and shared community structure detection.
网络嵌入的目的是学习保留结构信息的每个节点的潜在表示。许多现实世界的网络都有多个维度的节点和多种类型的关系。因此，用多路网络来表示这类网络更为合适。多路复用网络是由一组节点通过不同类型的链路连接在不同的层中形成的。然而，现有的基于随机行走的多路网络嵌入算法存在抽样偏差和关系类型不平衡的问题，导致其在下游任务中的性能较差。提出了一种基于自适应跨层森林火灾采样(FFS)的多路网络节点嵌入方法。我们首先关注FFS的抽样策略，以解决随机漫步的偏差问题。我们利用固定长度的队列来记录之前访问过的层，这可以平衡采样节点序列中不同层的边缘分布。此外，为了对节点的上下文进行自适应采样，我们还提出了一个节点的邻居划分系数(N P C)度量。节点序列的生成过程由NPC监督，用于自适应跨层采样。在不同领域的真实网络上进行的实验表明，我们的方法在跨域链接预测和共享社区结构检测等应用任务中优于最先进的方法。
{"title":"An Adaptive Cross-Layer Sampling-Based Node Embedding for Multiplex Networks","authors":"Nianwen Ning, Chenguang Song, Pengpeng Zhou, Yunlei Zhang, Bin Wu","doi":"10.1109/ICTAI.2019.00216","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00216","url":null,"abstract":"Network embedding aims to learn a latent representation of each node which preserves the structure information. Many real-world networks have multiple dimensions of nodes and multiple types of relations. Therefore, it is more appropriate to represent such kind of networks as multiplex networks. A multiplex network is formed by a set of nodes connected in different layers by links indicating interactions of different types. However, existing random walk based multiplex networks embedding algorithms have problems with sampling bias and imbalanced relation types, thus leading the poor performance in the downstream tasks. In this paper, we propose a node embedding method based on adaptive cross-layer forest fire sampling (FFS) for multiplex networks (FFME). We first focus on the sampling strategies of FFS to address the bias issue of random walk. We utilize a fixed-length queue to record previously visited layers, which can balance the edge distribution over different layers in sampled node sequences. In addition, to adaptively sample node's context, we also propose a metric for node called Neighbors Partition Coefficient (N P C ). The generation process of node sequence is supervised by NPC for adaptive cross-layer sampling. Experiments on real-world networks in diverse fields show that our method outperforms the state-of-the-art methods in application tasks such as cross-domain link prediction and shared community structure detection.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"83 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116408188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00218
Henok Ghebrechristos, G. Alaghband
We present a new system that can automatically generate input paths (syllabus) for a convolutional neural network to follow through a curriculum learning to improve training performance. Our system utilizes information-theoretic content measures of training samples to form syllabus at training time. We treat every sample as 2D random variable where a data point contained in the sample (such as a pixel) is modelled as an independent and identically distributed random variable (i.i.d) realization. We use several information theory methods to rank and determine when a sample is fed to a network by measuring its pixel composition and its relationship to other samples in the training set. Comparative evaluation of multiple state-of-the-art networks, including, GoogleNet, and VGG, on benchmark datasets demonstrate a syllabus that ranks samples using measures such as Joint Entropy between adjacent samples, can improve learning and significantly reduce the amount of training steps required to achieve desirable training accuracy. We present results that indicate our approach can reduce training loss by as much as a factor of 9 compared to conventional training.
{"title":"Optimizing Training using Information Theory-Based Curriculum Learning Factory","authors":"Henok Ghebrechristos, G. Alaghband","doi":"10.1109/ICTAI.2019.00218","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00218","url":null,"abstract":"We present a new system that can automatically generate input paths (syllabus) for a convolutional neural network to follow through a curriculum learning to improve training performance. Our system utilizes information-theoretic content measures of training samples to form syllabus at training time. We treat every sample as 2D random variable where a data point contained in the sample (such as a pixel) is modelled as an independent and identically distributed random variable (i.i.d) realization. We use several information theory methods to rank and determine when a sample is fed to a network by measuring its pixel composition and its relationship to other samples in the training set. Comparative evaluation of multiple state-of-the-art networks, including, GoogleNet, and VGG, on benchmark datasets demonstrate a syllabus that ranks samples using measures such as Joint Entropy between adjacent samples, can improve learning and significantly reduce the amount of training steps required to achieve desirable training accuracy. We present results that indicate our approach can reduce training loss by as much as a factor of 9 compared to conventional training.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115389060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00089
Xiuping Bao, Jiabin Yuan, Bei Chen
Human action recognition has became an important task in computer vision and has received a significant amount of research interests in recent years. Convolutional Neural Network (CNN) has shown its power in image recognition task. While in the field of video recognition, it is still a challenge problem. In this paper, we introduce a high-efficient attention-based convolutional network named ECPNet for video understanding. ECPNet adopts the framework that is a consecutive connection of 2D CNN and pseudo-3D CNN. The pseudo-3D means we replace the traditional 3 × 3 × 3 kernel with two 3D convolutional filters shaped 1 × 3 × 3 and 3 × 1 × 1. Our ECPNet combines the advantages of both 2D and 3D CNNs: (1) ECPNet is an end-to-end network and can learn information of appearance from images and motion between frames. (2) ECPNet requires less computing resource and lower memory consumption than many state-of-art models. (3) ECPNet is easy to expand for different demands of runtime and classification accuracy. We evaluate the proposed model on three popular video benchmarks in human action recognition task: Kinetics-mini (split of full Kinetics), UCF101 and HMDB51. Our ECPNet achieves the excellent performance on above datasets with less time cost.
{"title":"ECPNet: An Efficient Attention-Based Convolution Network with Pseudo-3D Block for Human Action Recognition","authors":"Xiuping Bao, Jiabin Yuan, Bei Chen","doi":"10.1109/ICTAI.2019.00089","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00089","url":null,"abstract":"Human action recognition has became an important task in computer vision and has received a significant amount of research interests in recent years. Convolutional Neural Network (CNN) has shown its power in image recognition task. While in the field of video recognition, it is still a challenge problem. In this paper, we introduce a high-efficient attention-based convolutional network named ECPNet for video understanding. ECPNet adopts the framework that is a consecutive connection of 2D CNN and pseudo-3D CNN. The pseudo-3D means we replace the traditional 3 × 3 × 3 kernel with two 3D convolutional filters shaped 1 × 3 × 3 and 3 × 1 × 1. Our ECPNet combines the advantages of both 2D and 3D CNNs: (1) ECPNet is an end-to-end network and can learn information of appearance from images and motion between frames. (2) ECPNet requires less computing resource and lower memory consumption than many state-of-art models. (3) ECPNet is easy to expand for different demands of runtime and classification accuracy. We evaluate the proposed model on three popular video benchmarks in human action recognition task: Kinetics-mini (split of full Kinetics), UCF101 and HMDB51. Our ECPNet achieves the excellent performance on above datasets with less time cost.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121404601","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00010
Sahil Verma, R. Yap
Symbolic execution is a powerful technique for bug finding and program testing. It is successful in finding bugs in real-world code. The core reasoning techniques use constraint solving, path exploration, and search, which are also the same techniques used in solving combinatorial problems, e.g., finite-domain constraint satisfaction problems (CSPs). We propose CSP instances as more challenging benchmarks to evaluate the effectiveness of the core techniques in symbolic execution. We transform CSP benchmarks into C programs suitable for testing the reasoning capabilities of symbolic execution tools. From a single CSP P, we transform P depending on transformation choice into different C programs. Preliminary testing with the KLEE, Tracer-X, and LLBMC tools show substantial runtime differences from transformation and solver choice. Our C benchmarks are effective in showing the limitations of existing symbolic execution tools. The motivation for this work is we believe that benchmarks of this form can spur the development and engineering of improved core reasoning in symbolic execution engines.
{"title":"Benchmarking Symbolic Execution Using Constraint Problems - Initial Results","authors":"Sahil Verma, R. Yap","doi":"10.1109/ICTAI.2019.00010","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00010","url":null,"abstract":"Symbolic execution is a powerful technique for bug finding and program testing. It is successful in finding bugs in real-world code. The core reasoning techniques use constraint solving, path exploration, and search, which are also the same techniques used in solving combinatorial problems, e.g., finite-domain constraint satisfaction problems (CSPs). We propose CSP instances as more challenging benchmarks to evaluate the effectiveness of the core techniques in symbolic execution. We transform CSP benchmarks into C programs suitable for testing the reasoning capabilities of symbolic execution tools. From a single CSP P, we transform P depending on transformation choice into different C programs. Preliminary testing with the KLEE, Tracer-X, and LLBMC tools show substantial runtime differences from transformation and solver choice. Our C benchmarks are effective in showing the limitations of existing symbolic execution tools. The motivation for this work is we believe that benchmarks of this form can spur the development and engineering of improved core reasoning in symbolic execution engines.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123310765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00244
Tianyang Wang, Jun Huan, Bo Li, Kaoning Hu
Real-world image denoising is a challenging but significant problem in computer vision. Unlike Gaussian denoising on which most existing methods focus, the real-world noise is nonadditive, and the distributions are difficult to model. This leads to unsatisfactory performance when applying a Gaussian denoising approach on real-world denoising problems. In this paper, we propose a simple framework for effective real-world image denoising. Specifically, we investigate the intrinsic properties of the Gaussian denoising prior and demonstrate this prior can aid real-world image denoising. To leverage this prior, we fine-tune it for only one epoch on a recently proposed real-world image denoising dataset, and the learned model can enhance both visual and quantitative results (peak-signal-noise-ratio) for real-world image denoising tasks. Extensive experiments demonstrate the effectiveness of our approach, and indicate that the Gaussian denoising prior can be also transferred to real-world image denoising by exploiting appropriate training schemes.
{"title":"Rethink Gaussian Denoising Prior for Real-World Image Denoising","authors":"Tianyang Wang, Jun Huan, Bo Li, Kaoning Hu","doi":"10.1109/ICTAI.2019.00244","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00244","url":null,"abstract":"Real-world image denoising is a challenging but significant problem in computer vision. Unlike Gaussian denoising on which most existing methods focus, the real-world noise is nonadditive, and the distributions are difficult to model. This leads to unsatisfactory performance when applying a Gaussian denoising approach on real-world denoising problems. In this paper, we propose a simple framework for effective real-world image denoising. Specifically, we investigate the intrinsic properties of the Gaussian denoising prior and demonstrate this prior can aid real-world image denoising. To leverage this prior, we fine-tune it for only one epoch on a recently proposed real-world image denoising dataset, and the learned model can enhance both visual and quantitative results (peak-signal-noise-ratio) for real-world image denoising tasks. Extensive experiments demonstrate the effectiveness of our approach, and indicate that the Gaussian denoising prior can be also transferred to real-world image denoising by exploiting appropriate training schemes.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125176766","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00213
Sheng Xie, Canlong Zhang, Zhixin Li, Zhiwen Wang
When extracting convolutional features from person images with low resolution, a large amount of available information will be lost due to the pooling, which will lead to the reduction of the accuracy of person classification models. This paper proposes a new classification model, which can effectively to reduce the loss of important information about the convolutional neural works. Firstly, the SE module in the Squeeze-and-Excitation Networks (SENet) is extracted and normalized to generate the Normalized Squeeze-and-Excitation (NSE) module. Then, 4 NSE modules are applied to the convolutional layers of ResNet. Finally, a Sparse Normalized Squeeze-and-Excitation Network (SNSENet) is constructed by adding 4 shortcut connections between the convolutional layers. The experimental results of Market-1501 show that the rank-1 of SNSE-ResNet-50 is 3.7% and 4.2% higher than that of SE-ResNet-50 and ResNet-50 respectively, it has done well in other person re-identification datasets.
{"title":"Sparse High-Level Attention Networks for Person Re-Identification","authors":"Sheng Xie, Canlong Zhang, Zhixin Li, Zhiwen Wang","doi":"10.1109/ICTAI.2019.00213","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00213","url":null,"abstract":"When extracting convolutional features from person images with low resolution, a large amount of available information will be lost due to the pooling, which will lead to the reduction of the accuracy of person classification models. This paper proposes a new classification model, which can effectively to reduce the loss of important information about the convolutional neural works. Firstly, the SE module in the Squeeze-and-Excitation Networks (SENet) is extracted and normalized to generate the Normalized Squeeze-and-Excitation (NSE) module. Then, 4 NSE modules are applied to the convolutional layers of ResNet. Finally, a Sparse Normalized Squeeze-and-Excitation Network (SNSENet) is constructed by adding 4 shortcut connections between the convolutional layers. The experimental results of Market-1501 show that the rank-1 of SNSE-ResNet-50 is 3.7% and 4.2% higher than that of SE-ResNet-50 and ResNet-50 respectively, it has done well in other person re-identification datasets.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122775297","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICTAI.2019.00079
Xiu Li, Guichun Duan, Zhouxia Wang, Jimmy S. J. Ren, Yongbing Zhang, Jiawei Zhang, Kaixiang Song
In the past a few years, we witnessed rapid advancement in face super-resolution from very low resolution(VLR) images. However, most of the previous studies focus on solving such problem without explicitly considering the impact of severe real-life image degradation (e.g. blur and noise). We can show that robustly recover details from VLR images is a task beyond the ability of current state-of-the-art method. In this paper, we borrow ideas from "facial composite" and propose an alternative approach to tackle this problem. We endow the degraded VLR images with additional cues by integrating existing face components from multiple reference images into a novel learning pipeline with both low level and high level semantic loss function as well as a specialized adversarial based training scheme. We show that our method is able to effectively and robustly restore relevant facial details from 16x16 images with extreme degradation. We also tested our approach against real-life images and our method performs favorably against previous methods.
{"title":"Recovering Extremely Degraded Faces by Joint Super-Resolution and Facial Composite","authors":"Xiu Li, Guichun Duan, Zhouxia Wang, Jimmy S. J. Ren, Yongbing Zhang, Jiawei Zhang, Kaixiang Song","doi":"10.1109/ICTAI.2019.00079","DOIUrl":"https://doi.org/10.1109/ICTAI.2019.00079","url":null,"abstract":"In the past a few years, we witnessed rapid advancement in face super-resolution from very low resolution(VLR) images. However, most of the previous studies focus on solving such problem without explicitly considering the impact of severe real-life image degradation (e.g. blur and noise). We can show that robustly recover details from VLR images is a task beyond the ability of current state-of-the-art method. In this paper, we borrow ideas from \"facial composite\" and propose an alternative approach to tackle this problem. We endow the degraded VLR images with additional cues by integrating existing face components from multiple reference images into a novel learning pipeline with both low level and high level semantic loss function as well as a specialized adversarial based training scheme. We show that our method is able to effectively and robustly restore relevant facial details from 16x16 images with extreme degradation. We also tested our approach against real-life images and our method performs favorably against previous methods.","PeriodicalId":346657,"journal":{"name":"2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128787724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：