IEICE Transactions on Information and Systems
Online ISSN : 1745-1361
Print ISSN : 0916-8532
Volume E105.D, Issue 2
Displaying 1-29 of 29 articles from this issue
Special Section on Blockchain Systems and Applications
  • Guojun WANG, Yizhi REN
    2022 Volume E105.D Issue 2 Pages 193-194
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS
    Download PDF (145K)
  • Hao WANG, Zhe LIU, Chunpeng GE, Kouichi SAKURAI, Chunhua SU
    Article type: INVITED PAPER
    2022 Volume E105.D Issue 2 Pages 195-204
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Smart contracts are becoming more and more popular in financial scenarios like medical insurance. Rather than traditional schemes, using smart contracts as a medium is a better choice for both participants, as it is fairer, more reliable, more efficient, and enables real-time payment. However, medical insurance contracts need to input the patient's condition information as the judgment logic to trigger subsequent execution. Since the blockchain is a closed network, it lacks a secure network environment for data interaction with the outside world. The Data feed aims to provide the service of the on-chain and off-chain data interaction. Existing researches on the data feed has solved the security problems on it effectively, such as Town Crier, TLS-N and they have also taken into account the privacy-preserving problems. However, these schemes cannot actually protect privacy because when the ciphertext data is executed by the contract, privacy information can still be inferred by analyzing the transaction results, since states of the contract are publicly visible. In this paper, based on zero-knowledge proof and Hawk technology, a on-and-off-chain complete smart contract data feed privacy-preserving scheme is proposed. In order to present our scheme more intuitively, we combined the medical insurance compensation case to implement it, which is called MIPDF. In our MIPDF, the patient and the insurance company are parties involved in the contract, and the hospital is the data provider of data feed. The patient's medical data is sent to the smart contract under the umbrella of the zero-knowledge proof signature scheme. The smart contract verifies the proof and calculates the insurance premium based on the judgment logic. Meanwhile, we use Hawk technology to ensure the privacy of on-chain contract execution, so that no information will be disclosed due to the result of contract execution. We give a general description of our scheme within the Universal Composability (UC) framework. We experiment and evaluate MIPDF on Ethereum for in-depth analysis. The results show that our scheme can securely and efficiently support the functions of medical insurance and achieve complete privacy-preserving.

    Download PDF (1968K)
  • Naoya OKANAMI, Ryuya NAKAMURA, Takashi NISHIDE
    Article type: INVITED PAPER
    2022 Volume E105.D Issue 2 Pages 205-214
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Sharding is a solution to the blockchain scalability problem. A sharded blockchain divides consensus nodes (validators) into groups called shards and processes transactions separately to improve throughput and latency. In this paper, we analyze the rational behavior of users in account/balance model-based sharded blockchains and identify a phenomenon in which accounts (users' wallets and smart contracts) eventually get concentrated in a few shards, making shard loads unfair. This phenomenon leads to bad user experiences, such as delays in transaction inclusions and increased transaction fees. To solve this problem, we propose two load balancing methods in account/balance model-based sharded blockchains. Both methods perform load balancing by periodically reassigning accounts: in the first method, the blockchain protocol itself performs load balancing and in the second method, wallets perform load balancing. We discuss the pros and cons of the two protocols, and apply the protocols to the execution sharding in Ethereum 2.0, an existing sharding design. Further, we analyze by simulation how the protocols behave to confirm that we can observe smaller transaction delays and fees. As a result, we released the simulation program as “Shargri-La,” a simulator designed for general-purpose user behavior analysis on the execution sharding in Ethereum 2.0.

    Download PDF (590K)
  • Tao PENG, Kejian GUAN, Jierong LIU
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 215-226
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    A mobile crowdsensing system (MCS) utilizes a crowd of users to collect large-scale data using their mobile devices efficiently. The collected data are usually linked with sensitive information, raising the concerns of user privacy leakage. To date, many approaches have been proposed to protect the users' privacy, with the majority relying on a centralized structure, which poses though attack and intrusion vulnerability. Some studies build a distributed platform exploiting a blockchain-type solution, which still requires a fully trusted third party (TTP) to manage a reliable reward distribution in the MCS. Spurred by the deficiencies of current methods, we propose a distributed user privacy protection structure that combines blockchain and a trusted execution environment (TEE). The proposed architecture successfully manages the users' privacy protection and an accurate reward distribution without requiring a TTP. This is because the encryption algorithms ensure data confidentiality and uncouple the correlation between the users' identity and the sensitive information in the collected data. Accordingly, the smart contract signature is used to manage the user deposit and verify the data. Extensive comparative experiments verify the efficiency and effectiveness of the proposed combined blockchain and TEE scheme.

    Download PDF (2099K)
  • Shingo FUJIMOTO, Takuma TAKEUCHI, Yoshiki HIGASHIKADO
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 227-234
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Blockchain is a distributed ledger technology used for trading digital assets, such as cryptocurrency, and trail records that need to be audited by third parties. The use cases of blockchain are expanding beyond cryptocurrency management. In particular, the token economy, in which tokenized assets are exchanged across different blockchain ledgers, is gaining popularity. Cross-chain technologies such as atomic swap have emerged as security technologies to realize this new use case of blockchain. However, existing approaches of cross-chain technology have unresolved issues, such as application limitations on different blockchain platforms owing to the incompatibility of the communication interface and crypto algorithm and inability to handle a complex business logic such as the escrow trade. In this study, the ConnectionChain is proposed, which enables the execution of an extended smart contract using abstracted operation on interworking ledgers. Moreover, field experimental results using the system prototype are presented and explained.

    Download PDF (1667K)
  • Haoyang AN, Jiageng CHEN
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 235-247
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    The development of educational informatization makes data privacy particularly important in education. With society's development, the education system is complicated, and the result of education evaluation becomes more and more critical to students. The evaluation process of education must be justice and transparent. In recent years, the Onscreen Marking (OSM) system based on traditional cloud platforms has been widely used in various large-scale public examinations. However, due to the excessive concentration of power in the existing scheme, the mainstream marking process is not transparent, and there are hidden dangers of black-box operation, which will damage the fairness of the examination. In addition, issues related to data security and privacy are still considered to be severe challenges. This paper deals with the above problems by providing secure and private transactions in a distributed OSM assuming the semi-trusted examination center. We have implemented a proof-of-concept for a consortium blockchain-based OSM in a privacy-preserving and auditable manner, enabling markers to mark on the distributed ledger anonymously. We have proposed a distributed OSM system in high-level, which provides theoretical support for the fair evaluation process of education informatization. It has particular theoretical and application value for education combined with blockchain.

    Download PDF (753K)
  • Takaki ASANUMA, Takanori ISOBE
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 248-255
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Hashcash, which is a Proof of Work (PoW) of bitcoin, is based on a preimage problem of hash functions of SHA-2 and RIPEMD. As these hash functions employ the Merkle-Damgard (MD) construction, a preimage can be found with negligible memory. Since such calculations can be accelerated by dedicated ASICs, it has a potential risk of a so-called 51% attack. To address this issue, we propose a new PoW scheme based on the key recovery problem of cascade block ciphers. By choosing the appropriate parameters, e.g., block sizes and key sizes of underlying block ciphers, we can make this problem a memory-hard problem such that it requires a lot of memory to efficiently solve it. Besides, we can independently adjust the required time complexity and memory complexity, according to requirements by target applications and progress of computational power.

    Download PDF (622K)
  • Zhuotao LIAN, Weizheng WANG, Huakun HUANG, Chunhua SU
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 256-263
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    In recent years, federated learning has attracted more and more attention as it could collaboratively train a global model without gathering the users' raw data. It has brought many challenges. In this paper, we proposed layer-based federated learning system with privacy preservation. We successfully reduced the communication cost by selecting several layers of the model to upload for global averaging and enhanced the privacy protection by applying local differential privacy. We evaluated our system in non independently and identically distributed scenario on three datasets. Compared with existing works, our solution achieved better performance in both model accuracy and training time.

    Download PDF (1111K)
  • Tong ZHANG, Yujue WANG, Yong DING, Qianhong WU, Hai LIANG, Huiyong WAN ...
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 264-271
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    With the development of Internet technology, the demand for signing electronic contracts has been greatly increased. The electronic contract generated by the participants in an online way enjoys the same legal effect as paper contract. The fairness is the key issue in jointly signing electronic contracts by the involved participants, so that all participants can either get the same copy of the contract or nothing. Most existing solutions only focus on the fairness of electronic contract generation between two participants, where the digital signature can effectively guarantee the fairness of the exchange of electronic contracts and becomes the conventional technology in designing the contract signing protocol. In this paper, an efficient blockchain-based multi-party electronic contract signing (MECS) protocol is presented, which not only offers the fairness of electronic contract generation for multiple participants, but also allows each participant to aggregate validate the signed copy of others. Security analysis shows that the proposed MECS protocol enjoys unforgeability, non-repudiation and fairness of electronic contracts, and performance analysis demonstrates the high efficiency of our construction.

    Download PDF (1691K)
  • Wenjuan LI, Yu WANG, Weizhi MENG, Jin LI, Chunhua SU
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 272-279
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    To safeguard critical services and assets in a distributed environment, collaborative intrusion detection systems (CIDSs) are usually adopted to share necessary data and information among various nodes, and enhance the detection capability. For simplifying the network management, software defined networking (SDN) is an emerging platform that decouples the controller plane from the data plane. Intuitively, SDN can help lighten the management complexity in CIDSs, and a CIDS can protect the security of SDN. In practical implementation, trust management is an important approach to help identify insider attacks (or malicious nodes) in CIDSs, but the challenge is how to ensure the data integrity when evaluating the reputation of a node. Motivated by the recent development of blockchain technology, in this work, we design BlockCSDN — a framework of blockchain-based collaborative intrusion detection in SDN, and take the challenge-based CIDS as a study. The experimental results under both external and internal attacks indicate that using blockchain technology can benefit the robustness and security of CIDSs and SDN.

    Download PDF (1073K)
  • Yingxiao XIANG, Chao LI, Tong CHEN, Yike LI, Endong TONG, Wenjia NIU, ...
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 280-288
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Controlled optimization of phases (COP) is a core implementation in the future intelligent traffic signal system (I-SIG), which has been deployed and tested in countries including the U.S. and China. In such a system design, optimal signal control depends on dynamic traffic situation awareness via connected vehicles. Unfortunately, I-SIG suffers data spoofing from any hacked vehicle; in particular, the spoofing of the last vehicle can break the system and cause severe traffic congestion. Specifically, coordinated attacks on multiple intersections may even bring cascading failure of the road traffic network. To mitigate this security issue, a blockchain-based multi-intersection joint defense mechanism upon COP planning is designed. The major contributions of this paper are the following. 1) A blockchain network constituted by road-side units at multiple intersections, which are originally distributed and decentralized, is proposed to obtain accurate and reliable spoofing detection. 2) COP-oriented smart contract is implemented and utilized to ensure the credibility of spoofing vehicle detection. Thus, an I-SIG can automatically execute a signal planning scheme according to traffic information without spoofing data. Security analysis for the data spoofing attack is carried out to demonstrate the security. Meanwhile, experiments on the simulation platform VISSIM and Hyperledger Fabric show the efficiency and practicality of the blockchain-based defense mechanism.

    Download PDF (2942K)
  • Yizhi REN, Zelong LI, Lifeng YUAN, Zhen ZHANG, Chunhua SU, Yujuan WANG ...
    Article type: PAPER
    2022 Volume E105.D Issue 2 Pages 289-299
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    The recommend system has been widely used in many web application areas such as e-commerce services. With the development of the recommend system, the HIN modeling method replaces the traditional bipartite graph modeling method to represent the recommend system. But several studies have already showed that recommend system is vulnerable to shilling attack (injecting attack). However, the effectiveness of how traditional shilling attack has rarely been studied directly in the HIN model. Moreover, no study has focused on how to enhance shilling attacks against HIN recommend system by using the high-level semantic information. This work analyzes the relationship between the high-level semantic information and the attacking effects in HIN recommend system. This work proves that attack results are proportional to the high-level semantic information. Therefore, we propose a heuristic attack method based on high-level semantic information, named Semantic Shilling Attack (SSA) on a HIN recommend system (HERec). This method injects a specific score into each selected item related to the target in semantics. It ensures transmitting the misleading information towards target items and normal users, and attempts to interfere with the effect of the recommend system. The experiment is dependent on two real-world datasets, and proves that the attacking effect is positively correlate with the number of meta-paths. The result shows that our method is more effective when compared with existing baseline algorithms.

    Download PDF (2139K)
  • Srinivas KOPPU, Kumar K, Siva Rama KRISHNAN SOMAYAJI, Iyapparaja MEENA ...
    Article type: SURVEY PAPER
    2022 Volume E105.D Issue 2 Pages 300-308
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Blockchain is one of the prominent rapidly used technology in the last decade in various applications. In recent years, many researchers explored the capabilities of blockchain in smart IoT to address various security challenges. Integration of IoT and blockchain solves the security problems but scalability still remains a huge challenge. To address this, various AI techniques can be applied in the blockchain IoT framework, thus providing an efficient information system. In this survey, various works pertaining to the domains which integrate AI, IoT and Blockchain has been explored. Also, this article discusses potential industrial use cases on fusion of blockchain, AI and IoT applications and its challenges.

    Download PDF (1223K)
Regular Section
  • Dhidhi PAMBUDI, Masaki KAWAMURA
    Article type: PAPER
    Subject area: Fundamentals of Information Systems
    2022 Volume E105.D Issue 2 Pages 309-319
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    We proposed a population-based metaheuristic called the spy algorithm for solving optimization problems and evaluated its performance. The design of our spy algorithm ensures the benefit of exploration and exploitation as well as cooperative and non-cooperative searches in each iteration. We compared the spy algorithm with genetic algorithm, improved harmony search, and particle swarm optimization on a set of non-convex functions that focus on accuracy, the ability of detecting many global optimum points, and computation time. From statistical analysis results, the spy algorithm outperformed the other algorithms. The spy algorithm had the best accuracy and detected more global optimum points within less computation time, indicating that our spy algorithm is more robust and faster then these other algorithms.

    Download PDF (2542K)
  • Yasutaka MATSUDA, Ryota SHIOYA, Hideki ANDO
    Article type: PAPER
    Subject area: Computer System
    2022 Volume E105.D Issue 2 Pages 320-332
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    The high energy consumption of current processors causes several problems, including a limited clock frequency, short battery lifetime, and reduced device reliability. It is therefore important to reduce the energy consumption of the processor. Among resources in a processor, the issue queue (IQ) is a large consumer of energy, much of which is consumed by the wakeup logic. Within the wakeup logic, the tag comparison that checks source operand readiness consumes a significant amount of energy. This paper proposes an energy reduction scheme for tag comparison, called double-stage tag comparison. This scheme first compares the lower bits of the tag and then, only if these match, compares the higher bits. Because the energy consumption of tag comparison is roughly proportional to the total number of bits compared, energy is saved by reducing this number. However, this sequential comparison increases the delay of the IQ, thereby increasing the clock cycle time. Although this can be avoided by allocating an extra cycle to the issue operation, this in turn degrades the IPC. To avoid IPC degradation, we reconfigure a small number of entries in the IQ, where several oldest instructions that are likely to have an adverse effect on performance reside, to a single stage for tag comparison. Our evaluation results for SPEC2017 benchmark programs show that the double-stage tag comparison achieves on average a 21% reduction in the energy consumed by the wakeup logic (15% when including the overhead) with only 3.0% performance degradation.

    Download PDF (1158K)
  • Ryunosuke NAGAYAMA, Ryohei BANNO, Kazuyuki SHUDO
    Article type: PAPER
    Subject area: Data Engineering, Web Information Systems
    2022 Volume E105.D Issue 2 Pages 333-343
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    In Bitcoin and Ethereum, nodes require a large storage capacity to maintain all of the blockchain data such as transactions. As of September 2021, the storage size of the Bitcoin blockchain has expanded to 355 GB, and it has increased by approximately 50 GB every year over the last five years. This storage requirement is a major hurdle to becoming a block proposer or validator. We propose an architecture called Trail that allows nodes to hold all blocks in a small storage and to generate and validate blocks and transactions. A node in Trail holds all blocks without transactions, UTXOs or account balances. The block size is approximately 8 kB, which is 100 times smaller than that of Bitcoin. On the other hand, a client who issues transactions needs to hold proof of its assets. Thus, compared to traditional blockchains, clients must store additional data. We show that proper data archiving can keep the account device storage size small. Then, we propose a method of executing smart contracts in Trail using a threshold signature. Trail allows more users to be block proposers and validators and improves the decentralization and security of the blockchain.

    Download PDF (682K)
  • Meiming FU, Qingyang LIU, Jiayi LIU, Xiang WANG, Hongyan YANG
    Article type: PAPER
    Subject area: Information Network
    2022 Volume E105.D Issue 2 Pages 344-354
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Network virtualization has become a promising paradigm for supporting diverse vertical services in Software Defined Networks (SDNs). Each vertical service is carried by a virtual network (VN), which normally has a chaining structure. In this way, a Service Function Chain (SFC) is composed by an ordered set of virtual network functions (VNFs) to provide tailored network services. Such new programmable flexibilities for future networks also bring new network management challenges: how to collect and analyze network measurement data, and further predict and diagnose the performance of SFCs? This is a fundamental problem for the management of SFCs, because the VNFs could be migrated in case of SFC performance degradation to avoid Service Level Agreement (SLA) violation. Despite the importance of the problem, SFC performance analysis has not attracted much research attention in the literature. In this current paper, enabled by a novel detailed network debugging technology, In-band Network Telemetry (INT), we propose a learning based framework for early SFC fault prediction and diagnosis. Based on the SFC traffic flow measurement data provided by INT, the framework firstly extracts SFC performance features. Then, Long Short-Term Memory (LSTM) networks are utilized to predict the upcoming values for these features in the next time slot. Finally, Support Vector Machine (SVM) is utilized as network fault classifier to predict possible SFC faults. We also discuss the practical utilization relevance of the proposed framework, and conduct a set of network emulations to validate the performance of the proposed framework.

    Download PDF (3927K)
  • Jianyong DUAN, Liangcai LI, Mei ZHANG, Hao WANG
    Article type: PAPER
    Subject area: Artificial Intelligence, Data Mining
    2022 Volume E105.D Issue 2 Pages 355-363
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Personalized news recommendation is becoming increasingly important for online news platforms to help users alleviate information overload and improve news reading experience. A key problem in news recommendation is learning accurate user representations to capture their interest. However, most existing news recommendation methods usually learn user representation only from their interacted historical news, while ignoring the clustering features among users. Here we proposed a hierarchical user preference hash network to enhance the representation of users' interest. In the hash part, a series of buckets are generated based on users' historical interactions. Users with similar preferences are assigned into the same buckets automatically. We also learn representations of users from their browsed news in history part. And then, a Route Attention is adopted to combine these two parts (history vector and hash vector) and get the more informative user preference vector. As for news representation, a modified transformer with category embedding is exploited to build news semantic representation. By comparing the hierarchical hash network with multiple news recommendation methods and conducting various experiments on the Microsoft News Dataset (MIND) validate the effectiveness of our approach on news recommendation.

    Download PDF (1513K)
  • Hiroki ISHIGURO, Takashi ISHIDA, Masashi SUGIYAMA
    Article type: PAPER
    Subject area: Artificial Intelligence, Data Mining
    2022 Volume E105.D Issue 2 Pages 364-376
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    It has been demonstrated that large-scale labeled datasets facilitate the success of machine learning. However, collecting labeled data is often very costly and error-prone in practice. To cope with this problem, previous studies have considered the use of a complementary label, which specifies a class that an instance does not belong to and can be collected more easily than ordinary labels. However, complementary labels could also be error-prone and thus mitigating the influence of label noise is an important challenge to make complementary-label learning more useful in practice. In this paper, we derive conditions for the loss function such that the learning algorithm is not affected by noise in complementary labels. Experiments on benchmark datasets with noisy complementary labels demonstrate that the loss functions that satisfy our conditions significantly improve the classification performance.

    Download PDF (1166K)
  • Shuhei YAMAMOTO, Takeshi KURASHIMA, Hiroyuki TODA
    Article type: PAPER
    Subject area: Artificial Intelligence, Data Mining
    2022 Volume E105.D Issue 2 Pages 377-386
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Front video and sensor data captured by vehicle-mounted event recorders are used for not only traffic accident evidence but also safe-driving education as near-miss traffic incident data. However, most event recorder (ER) data shows only regular driving events. To utilize near-miss data for safe-driving education, we need to be able to easily and rapidly locate the appropriate data from large amounts of ER data through labels attached to the scenes/events of interest. This paper proposes a method that can automatically identify near-misses with objects such as pedestrians and bicycles by processing the ER data. The proposed method extracts two deep feature representations that consider car status and the environment surrounding the car. The first feature representation is generated by considering the temporal transitions of car status. The second one can extract the positional relationship between the car and surrounding objects by processing object detection results. Experiments on actual ER data demonstrate that the proposed method can accurately identify and tag near-miss events.

    Download PDF (1373K)
  • Yuichiro NOMURA, Takio KURITA
    Article type: PAPER
    Subject area: Artificial Intelligence, Data Mining
    2022 Volume E105.D Issue 2 Pages 387-395
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    In the recent years, deep learning has achieved significant results in various areas of machine learning. Deep learning requires a huge amount of data to train a model, and data collection techniques such as web crawling have been developed. However, there is a risk that these data collection techniques may generate incorrect labels. If a deep learning model for image classification is trained on a dataset with noisy labels, the generalization performance significantly decreases. This problem is called Learning with Noisy Labels (LNL). One of the recent researches on LNL, called DivideMix [1], has successfully divided the dataset into samples with clean labels and ones with noisy labels by modeling loss distribution of all training samples with a two-component Mixture Gaussian model (GMM). Then it treats the divided dataset as labeled and unlabeled samples and trains the classification model in a semi-supervised manner. Since the selected samples have lower loss values and are easy to classify, training models are in a risk of overfitting to the simple pattern during training. To train the classification model without overfitting to the simple patterns, we propose to introduce consistency regularization on the selected samples by GMM. The consistency regularization perturbs input images and encourages model to outputs the same value to the perturbed images and the original images. The classification model simultaneously receives the samples selected as clean and their perturbed ones, and it achieves higher generalization performance with less overfitting to the selected samples. We evaluated our method with synthetically generated noisy labels on CIFAR-10 and CIFAR-100 and obtained results that are comparable or better than the state-of-the-art method.

    Download PDF (976K)
  • Masayuki HIROMOTO, Hisanao AKIMA, Teruo ISHIHARA, Takuji YAMAMOTO
    Article type: PAPER
    Subject area: Pattern Recognition
    2022 Volume E105.D Issue 2 Pages 396-405
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Zero-shot learning (ZSL) aims to classify images of unseen classes by learning relationship between visual and semantic features. Existing works have been improving recognition accuracy from various approaches, but they employ computationally intensive algorithms that require iterative optimization. In this work, we revisit the primary approach of the pattern recognition, ı.e., nearest neighbor classifiers, to solve the ZSL task by an extremely simple and fast way, called SimpleZSL. Our algorithm consists of the following three simple techniques: (1) just averaging feature vectors to obtain visual prototypes of seen classes, (2) calculating a pseudo-inverse matrix via singular value decomposition to generate visual features of unseen classes, and (3) inferring unseen classes by a nearest neighbor classifier in which cosine similarity is used to measure distance between feature vectors. Through the experiments on common datasets, the proposed method achieves good recognition accuracy with drastically small computational costs. The execution time of the proposed method on a single CPU is more than 100 times faster than those of the GPU implementations of the existing methods with comparable accuracies.

    Download PDF (726K)
  • Masayuki MIYAMA
    Article type: PAPER
    Subject area: Image Recognition, Computer Vision
    2022 Volume E105.D Issue 2 Pages 406-414
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Object contour detection is a task of extracting the shape created by the boundaries between objects in an image. Conventional methods limit the detection targets to specific categories, or miss-detect edges of patterns inside an object. We propose a new method to represent a contour image where the pixel value is the distance to the boundary. Contour detection becomes a regression problem that estimates this contour image. A deep convolutional network for contour estimation is combined with stereo vision to detect unspecified object contours. Furthermore, thanks to similar inference targets and common network structure, we propose a network that simultaneously estimates both contour and disparity with fully shared weights. As a result of experiments, the multi-tasking network drew a good precision-recall curve, and F-measure was about 0.833 for FlyingThings3D dataset. L1 loss of disparity estimation for the dataset was 2.571. This network reduces the amount of calculation and memory capacity by half, and accuracy drop compared to the dedicated networks is slight. Then we quantize both weights and activations of the network to 3-bit. We devise a dedicated hardware architecture for the quantized CNN and implement it on an FPGA. This circuit uses only internal memory to perform forward propagation calculations, that eliminates high-power external memory accesses. This circuit is a stall-free pixel-by-pixel pipeline, and performs 8 rows, 16 input channels, 16 output channels, 3 by 3 pixels convolution calculations in parallel. The convolution calculation performance at the operating frequency of 250 MHz is 9 TOPs/s.

    Download PDF (1373K)
  • Masashi NISHIYAMA, Michiko INOUE, Yoshio IWAI
    Article type: PAPER
    Subject area: Image Recognition, Computer Vision
    2022 Volume E105.D Issue 2 Pages 415-426
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    We propose an attention mechanism in deep learning networks for gender recognition using the gaze distribution of human observers when they judge the gender of people in pedestrian images. Prevalent attention mechanisms spatially compute the correlation among values of all cells in an input feature map to calculate attention weights. If a large bias in the background of pedestrian images (e.g., test samples and training samples containing different backgrounds) is present, the attention weights learned using the prevalent attention mechanisms are affected by the bias, which in turn reduces the accuracy of gender recognition. To avoid this problem, we incorporate an attention mechanism called gaze-guided self-attention (GSA) that is inspired by human visual attention. Our method assigns spatially suitable attention weights to each input feature map using the gaze distribution of human observers. In particular, GSA yields promising results even when using training samples with the background bias. The results of experiments on publicly available datasets confirm that our GSA, using the gaze distribution, is more accurate in gender recognition than currently available attention-based methods in the case of background bias between training and test samples.

    Download PDF (2038K)
  • Young H. OH, Yunho JIN, Tae Jun HAM, Jae W. LEE
    Article type: LETTER
    Subject area: Fundamentals of Information Systems
    2022 Volume E105.D Issue 2 Pages 427-431
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Many cloud service providers employ specialized hardware accelerators, called neural processing units (NPUs), to accelerate deep neural networks (DNNs). An NPU scheduler is responsible for scheduling incoming user requests and required to satisfy the two, often conflicting, optimization goals: maximizing system throughput and satisfying quality-of-service (QoS) constraints (e.g., deadlines) of individual requests. We propose Layerweaver+, a low-cost layer-wise DNN scheduler for NPUs, which provides both high system throughput and minimal QoS violations. For a serving scenario based on the industry-standard MLPerf inference benchmark, Layerweaver+ significantly improves the system throughput by up to 266.7% over the baseline scheduler serving one DNN at a time.

    Download PDF (457K)
  • Fei WU, Xinhao ZHENG, Ying SUN, Yang GAO, Xiao-Yuan JING
    Article type: LETTER
    Subject area: Software Engineering
    2022 Volume E105.D Issue 2 Pages 432-435
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Cross-project defect prediction (CPDP) is a hot research topic in recent years. The inconsistent data distribution between source and target projects and lack of labels for most of target instances bring a challenge for defect prediction. Researchers have developed several CPDP methods. However, the prediction performance still needs to be improved. In this paper, we propose a novel approach called Joint Domain Adaption and Pseudo-Labeling (JDAPL). The network architecture consists of a feature mapping sub-network to map source and target instances into a common subspace, followed by a classification sub-network and an auxiliary classification sub-network. The classification sub-network makes use of the label information of labeled instances to generate pseudo-labels. The auxiliary classification sub-network learns to reduce the distribution difference and improve the accuracy of pseudo-labels for unlabeled instances through loss maximization. Network training is guided by the adversarial scheme. Extensive experiments are conducted on 10 projects of the AEEEM and NASA datasets, and the results indicate that our approach achieves better performance compared with the baselines.

    Download PDF (2885K)
  • Kazuki SATO, Satoshi NAKATA, Takashi MATSUBARA, Kuniaki UEHARA
    Article type: LETTER
    Subject area: Pattern Recognition
    2022 Volume E105.D Issue 2 Pages 436-440
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    There exists a great demand for automatic anomaly detection in industrial world. The anomaly has been defined as a group of samples that rarely or never appears. Given a type of products, one has to collect numerous samples and train an anomaly detector. When one diverts a model trained with old types of products with sufficient inventory to the new type, one can detect anomalies of the new type before a production line is established. However, because of the definition of the anomaly, a typical anomaly detector considers the new type of products anomalous even if it is consistent with the standard. Given the above practical demand, this study propose a novel problem setting, few-shot anomaly detection, where an anomaly detector trained in source domains is adapted to a small set of target samples without full retraining. Then, we tackle this problem using a hierarchical probabilistic model based on deep learning. Our empirical results on toy and real-world datasets demonstrate that the proposed model detects anomalies in a small set of target samples successfully.

    Download PDF (1270K)
  • Jie ZHU, Yuan ZONG, Hongli CHANG, Li ZHAO, Chuangao TANG
    Article type: LETTER
    Subject area: Image Recognition, Computer Vision
    2022 Volume E105.D Issue 2 Pages 441-445
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Unsupervised domain adaptation (DA) is a challenging machine learning problem since the labeled training (source) and unlabeled testing (target) sets belong to different domains and then have different feature distributions, which has recently attracted wide attention in micro-expression recognition (MER). Although some well-performing unsupervised DA methods have been proposed, these methods cannot well solve the problem of unsupervised DA in MER, a. k. a., cross-domain MER. To deal with such a challenging problem, in this letter we propose a novel unsupervised DA method called Joint Patch weighting and Moment Matching (JPMM). JPMM bridges the source and target micro-expression feature sets by minimizing their probability distribution divergence with a multi-order moment matching operation. Meanwhile, it takes advantage of the contributive facial patches by the weight learning such that a domain-invariant feature representation involving micro-expression distinguishable information can be learned. Finally, we carry out extensive experiments to evaluate the proposed JPMM method is superior to recent state-of-the-art unsupervised DA methods in dealing with cross-domain MER.

    Download PDF (151K)
  • Fairuz Safwan MAHAD, Masakazu IWAMURA, Koichi KISE
    Article type: LETTER
    Subject area: Image Recognition, Computer Vision
    2022 Volume E105.D Issue 2 Pages 446-449
    Published: February 01, 2022
    Released on J-STAGE: February 01, 2022
    JOURNAL FREE ACCESS

    Neural network-based three-dimensional (3D) reconstruction methods have produced promising results. However, they do not pay particular attention to reconstructing detailed parts of objects. This occurs because the network is not designed to capture the fine details of objects. In this paper, we propose a network designed to capture both the coarse and fine details of objects to improve the reconstruction of the fine parts of objects.

    Download PDF (505K)
feedback
Top