Takashi Obi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Takashi Obi has a total of 81 co-authors. Here are the top co-authors by number of shared publications:
- Masahiro Yamaguchi
Institute of Science Tokyo, Kanagawa, Japan
16 shared publications Nagaaki Ohyama
Tokyo Institute of Technology
13 shared publications- Hiroyuki Suzuki
Institute of Innovative Research, Tokyo Institute of Technology, Yokohama, Japan
7 shared publications - Joong-Sun Lee
Institute of Innovative Research, Tokyo Institute of Technology, Yokohama, Japan
7 shared publications - Dara Tith
Institute of Innovative Research, Tokyo Institute of Technology, Yokohama, Japan
4 shared publications - Eleni Aloupogianni
Interactive Coventry Ltd., Coventry, U.K.
4 shared publications
Latest Publications
- Open Access
- 9 October 2025
- Published by Elsevier in Internet of Things
- Vol. 34
DIDAuth-IoTFW: Decentralized firmware authentication for smart home IoT devices using verifiable credentials
- 9 October 2025
- Published by Elsevier in Internet of Things
- Vol. 34
Open AccessDIDAuth-IoTFW: Decentralized firmware authentication for smart home IoT devices using verifiable credentials
- Open Access
- 1 April 2025
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 26
FocusAugMix: A data augmentation method for enhancing Acute Lymphoblastic Leukemia classification
- 1 April 2025
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 26
Open AccessFocusAugMix: A data augmentation method for enhancing Acute Lymphoblastic Leukemia classification
- Open Access
- 8 April 2024
- Published by Springer Nature in Machine Vision and Applications
- Vol. 35 (3) , 1-13
The improvement of ground truth annotation in public datasets for human detection
- 8 April 2024
- Published by Springer Nature in Machine Vision and Applications
- Vol. 35 (3) , 1-13
Open AccessThe improvement of ground truth annotation in public datasets for human detection
- Open Access
- 7 March 2024
- Published by Springer Nature in International Journal of Information Security
- Vol. 23 (3) , 1927-1937
Security-enhanced firmware management scheme for smart home IoT devices using distributed ledger technologies
With the increase of IoT devices generating large amounts of user-sensitive data, improper firmware harms users’ security and privacy. Latest home appliances are integrated with features to assure compatibility with smart home IoT. However, applying complex security mechanisms to IoT is limited by device hardware capabilities, making them vulnerable to attacks. Such attacks have recently become frequent. To address this issue, we developed a secure verification mechanism for firmware released by the device’s manufacturer. We proposed an IoT gateway for secure firmware verification and updating for smart home IoT devices utilizing the IOTA MAM (Masked Authenticated Messaging) protocol and a distributed file system with IPFS (Inter-Planetary File System) protocol. These two communication protocols ensure decentralized communication and firmware file distribution between the IoT device vendor and the IoT end device. The proposed scheme securely shares latest firmware content over IOTA and IPFS networks, performs a secure firmware update on IoT end devices and ensures authenticity and integrity of the firmware. Two types of validation methods were proposed for firmware updating and validation. We implemented the proposed scheme using three entities, Vendor, IoT gateway, and IoT end device. Our system yielded promising results in performing secure automated firmware updates on IoT end devices with very low computational power. The system’s functionality was implemented using IOTA’s MAM run on Raspberry Pi as an IoT gateway along with an ESP8266 Wi-Fi microcontroller, demonstrating the effectiveness of our approach. Our proposed methodology can be used for secure firmware distribution on home IoT applications.
- 7 March 2024
- Published by Springer Nature in International Journal of Information Security
- Vol. 23 (3) , 1927-1937
Open AccessSecurity-enhanced firmware management scheme for smart home IoT devices using distributed ledger technologies
With the increase of IoT devices generating large amounts of user-sensitive data, improper firmware harms users’ security and privacy. Latest home appliances are integrated with features to assure compatibility with smart home IoT. However, applying complex security mechanisms to IoT is limited by device hardware capabilities, making them vulnerable to attacks. Such attacks have recently become frequent. To address this issue, we developed a secure verification mechanism for firmware released by the device’s manufacturer. We proposed an IoT gateway for secure firmware verification and updating for smart home IoT devices utilizing the IOTA MAM (Masked Authenticated Messaging) protocol and a distributed file system with IPFS (Inter-Planetary File System) protocol. These two communication protocols ensure decentralized communication and firmware file distribution between the IoT device vendor and the IoT end device. The proposed scheme securely shares latest firmware content over IOTA and IPFS networks, performs a secure firmware update on IoT end devices and ensures authenticity and integrity of the firmware. Two types of validation methods were proposed for firmware updating and validation. We implemented the proposed scheme using three entities, Vendor, IoT gateway, and IoT end device. Our system yielded promising results in performing secure automated firmware updates on IoT end devices with very low computational power. The system’s functionality was implemented using IOTA’s MAM run on Raspberry Pi as an IoT gateway along with an ESP8266 Wi-Fi microcontroller, demonstrating the effectiveness of our approach. Our proposed methodology can be used for secure firmware distribution on home IoT applications.
- Open Access
- 2 January 2024
- Published by Springer Nature in International Journal of Information Technology
- Vol. 16 (3) , 1279-1292
Integrating prior knowledge to build transformer models
The big Artificial General Intelligence models inspire hot topics currently. The black box problems of Artificial Intelligence (AI) models still exist and need to be solved urgently, especially in the medical area. Therefore, transparent and reliable AI models with small data are also urgently necessary. To build a trustable AI model with small data, we proposed a prior knowledge-integrated transformer model. We first acquired prior knowledge using Shapley Additive exPlanations from various pre-trained machine learning models. Then, we used the prior knowledge to construct the transformer models and compared our proposed models with the Feature Tokenization Transformer model and other classification models. We tested our proposed model on three open datasets and one non-open public dataset in Japan to confirm the feasibility of our proposed methodology. Our results certified that knowledge-integrated transformer models perform better (1%) than general transformer models. Meanwhile, our proposed methodology identified that the self-attention of factors in our proposed transformer models is nearly the same, which needs to be explored in future work. Moreover, our research inspires future endeavors in exploring transparent small AI models.
- 2 January 2024
- Published by Springer Nature in International Journal of Information Technology
- Vol. 16 (3) , 1279-1292
Open AccessIntegrating prior knowledge to build transformer models
The big Artificial General Intelligence models inspire hot topics currently. The black box problems of Artificial Intelligence (AI) models still exist and need to be solved urgently, especially in the medical area. Therefore, transparent and reliable AI models with small data are also urgently necessary. To build a trustable AI model with small data, we proposed a prior knowledge-integrated transformer model. We first acquired prior knowledge using Shapley Additive exPlanations from various pre-trained machine learning models. Then, we used the prior knowledge to construct the transformer models and compared our proposed models with the Feature Tokenization Transformer model and other classification models. We tested our proposed model on three open datasets and one non-open public dataset in Japan to confirm the feasibility of our proposed methodology. Our results certified that knowledge-integrated transformer models perform better (1%) than general transformer models. Meanwhile, our proposed methodology identified that the self-attention of factors in our proposed transformer models is nearly the same, which needs to be explored in future work. Moreover, our research inspires future endeavors in exploring transparent small AI models.
- Open Access
- 1 May 2023
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 28 (05) , 056501-18
Generative adversarial network based digital stain conversion for generating RGB EVG stained image from hyperspectral H&E stained image
SignificanceQuantification of elastic fiber in the tissue specimen is an important aspect of diagnosing different diseases. Though hematoxylin and eosin (H&E) staining is a routinely used and less expensive tissue staining technique, elastic and collagen fibers cannot be differentiated using it. So, in conventional pathology, special staining technique, such as Verhoeff’s van Gieson (EVG), is applied physically for this purpose. However, the procedure of EVG staining is very expensive and time-consuming.AimThe goal of our study is to propose a deep-learning-based computerized method for the generation of RGB EVG stained tissue from hyperspectral H&E stained one to save the time and cost of conventional EVG staining procedure.ApproachH&E stained hyperspectral image and EVG stained RGB whole slide image of human pancreatic tissue have been leveraged for this experiment. CycleGAN-based deep learning model has been proposed for digital stain conversion while images from source and target domains are of different modalities (hyperspectral and RGB) with different channel dimensions. A set of three basis functions have been introduced for calculating one of the losses of the proposed method, which retains the relevant features of EVG stained image within the reduced channel dimension of the H&E stained one.ResultsThe experimental results showed that a set of three basis functions including linear discriminant function and transmittance spectrum of eosin and hematoxylin better retained the essential properties of the elastic fiber to be discriminated from collagen fiber within the reduced dimension of the hyperspectral H&E stained image. Also, only a smaller number of paired training data with our proposed training method contributed significantly to the generation of more realistic EVG stained image with more precise identification of elastic fiber.ConclusionsRGB EVG stained image is generated from hyperspectral H&E stained image for which our model has performed two types of image conversion simultaneously: hyperspectral to RGB and H&E to EVG. The experimental results show that the intentionally designed set of three basis functions contains more relevant information and prove the effectiveness of our proposed method in generating realistic RGB EVG stained image from hyperspectral H&E stained one.
- 1 May 2023
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 28 (05) , 056501-18
Open AccessGenerative adversarial network based digital stain conversion for generating RGB EVG stained image from hyperspectral H&E stained image
SignificanceQuantification of elastic fiber in the tissue specimen is an important aspect of diagnosing different diseases. Though hematoxylin and eosin (H&E) staining is a routinely used and less expensive tissue staining technique, elastic and collagen fibers cannot be differentiated using it. So, in conventional pathology, special staining technique, such as Verhoeff’s van Gieson (EVG), is applied physically for this purpose. However, the procedure of EVG staining is very expensive and time-consuming.AimThe goal of our study is to propose a deep-learning-based computerized method for the generation of RGB EVG stained tissue from hyperspectral H&E stained one to save the time and cost of conventional EVG staining procedure.ApproachH&E stained hyperspectral image and EVG stained RGB whole slide image of human pancreatic tissue have been leveraged for this experiment. CycleGAN-based deep learning model has been proposed for digital stain conversion while images from source and target domains are of different modalities (hyperspectral and RGB) with different channel dimensions. A set of three basis functions have been introduced for calculating one of the losses of the proposed method, which retains the relevant features of EVG stained image within the reduced channel dimension of the H&E stained one.ResultsThe experimental results showed that a set of three basis functions including linear discriminant function and transmittance spectrum of eosin and hematoxylin better retained the essential properties of the elastic fiber to be discriminated from collagen fiber within the reduced dimension of the hyperspectral H&E stained image. Also, only a smaller number of paired training data with our proposed training method contributed significantly to the generation of more realistic EVG stained image with more precise identification of elastic fiber.ConclusionsRGB EVG stained image is generated from hyperspectral H&E stained image for which our model has performed two types of image conversion simultaneously: hyperspectral to RGB and H&E to EVG. The experimental results show that the intentionally designed set of three basis functions contains more relevant information and prove the effectiveness of our proposed method in generating realistic RGB EVG stained image from hyperspectral H&E stained one.
- Open Access
- 29 April 2023
- Published by Springer Nature in International Journal of Information Technology
- Vol. 15 (4) , 1759-1768
XAI-based cross-ensemble feature ranking methodology for machine learning models
Artificial Intelligence (AI) as one robust technology has been used in various fields, making innovative society possible and changing our lifestyles. However, the black box problem is still one big problem for artificial intelligence. In this study, we first compared the results of kernel Shapley Additive exPlanations (SHAP) for various machine learning models and found that the single SHAP model cannot explain the models at the human knowledge level. Then the factors’ global ranking was calculated using our proposed ensemble methodology. Finally, the new factors’ ranking was compared with other factor ranking method. Our experimental results declare that the proposed cross-ensemble feature ranking methodology provides stable and comparatively reliable feature ranking in both the classification and regression models.
- 29 April 2023
- Published by Springer Nature in International Journal of Information Technology
- Vol. 15 (4) , 1759-1768
Open AccessXAI-based cross-ensemble feature ranking methodology for machine learning models
Artificial Intelligence (AI) as one robust technology has been used in various fields, making innovative society possible and changing our lifestyles. However, the black box problem is still one big problem for artificial intelligence. In this study, we first compared the results of kernel Shapley Additive exPlanations (SHAP) for various machine learning models and found that the single SHAP model cannot explain the models at the human knowledge level. Then the factors’ global ranking was calculated using our proposed ensemble methodology. Finally, the new factors’ ranking was compared with other factor ranking method. Our experimental results declare that the proposed cross-ensemble feature ranking methodology provides stable and comparatively reliable feature ranking in both the classification and regression models.
- Open Access
- 26 January 2023
- Published by Springer Nature in Health and Technology
- Vol. 13 (1) , 119-131
Interpretable machine learning analysis to identify risk factors for diabetes using the anonymous living census data of Japan
Purpose: Diabetes mellitus causes various problems in our life. With the big data boom in our society, some risk factors for Diabetes must still exist. To identify new risk factors for diabetes in the big data society and explore further efficient use of big data, the non-objective-oriented census data about the Japanese Citizen’s Survey of Living Conditions were analyzed using interpretable machine learning methods. Methods: Seven interpretable machine learning methods were used to analysis Japan citizens’ census data. Firstly, logistic analysis was used to analyze the risk factors of diabetes from 19 selected initial elements. Then, the linear analysis, linear discriminate analysis, Hayashi’s quantification analysis method 2, random forest, XGBoost, and SHAP methods were used to re-check and find the different factor contributions. Finally, the relationship among the factors was analyzed to understand the relationship among factors. Results: Four new risk factors: the number of family members, insurance type, public pension type, and health awareness level, were found as risk factors for diabetes mellitus for the first time, while another 11 risk factors were reconfirmed in this analysis. Especially the insurance type factor and health awareness level factor make more contributions to diabetes than factors: hypertension, hyperlipidemia, and stress in some interpretable models. We also found that work years were identified as a risk factor for diabetes because it has a high coefficient with the risk factor of age. Conclusions: New risk factors for diabetes mellitus were identified based on Japan's non-objective-oriented anonymous census data using interpretable machine learning models. The newly identified risk factors inspire new possible policies for preventing diabetes. Moreover, our analysis certifies that big data can help us find helpful knowledge in today's prosperous society. Our study also paves the way for identifying more risk factors and promoting the efficiency of using big data.
- 26 January 2023
- Published by Springer Nature in Health and Technology
- Vol. 13 (1) , 119-131
Open AccessInterpretable machine learning analysis to identify risk factors for diabetes using the anonymous living census data of Japan
Purpose: Diabetes mellitus causes various problems in our life. With the big data boom in our society, some risk factors for Diabetes must still exist. To identify new risk factors for diabetes in the big data society and explore further efficient use of big data, the non-objective-oriented census data about the Japanese Citizen’s Survey of Living Conditions were analyzed using interpretable machine learning methods. Methods: Seven interpretable machine learning methods were used to analysis Japan citizens’ census data. Firstly, logistic analysis was used to analyze the risk factors of diabetes from 19 selected initial elements. Then, the linear analysis, linear discriminate analysis, Hayashi’s quantification analysis method 2, random forest, XGBoost, and SHAP methods were used to re-check and find the different factor contributions. Finally, the relationship among the factors was analyzed to understand the relationship among factors. Results: Four new risk factors: the number of family members, insurance type, public pension type, and health awareness level, were found as risk factors for diabetes mellitus for the first time, while another 11 risk factors were reconfirmed in this analysis. Especially the insurance type factor and health awareness level factor make more contributions to diabetes than factors: hypertension, hyperlipidemia, and stress in some interpretable models. We also found that work years were identified as a risk factor for diabetes because it has a high coefficient with the risk factor of age. Conclusions: New risk factors for diabetes mellitus were identified based on Japan's non-objective-oriented anonymous census data using interpretable machine learning models. The newly identified risk factors inspire new possible policies for preventing diabetes. Moreover, our analysis certifies that big data can help us find helpful knowledge in today's prosperous society. Our study also paves the way for identifying more risk factors and promoting the efficiency of using big data.
- Open Access
- 31 October 2022
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 27 (10) , 106007-22
Hyperspectral imaging for tumor segmentation on pigmented skin lesions
In Japan, incidence and mortality rates of skin cancer have increased in the past decades, despite adjustments for an aging population.1 At the same time, the American Cancer Society estimates that malignant skin tumors comprise the majority of all cancers in the United States.2 Early diagnosis is key for the effective treatment of skin tumors. Dermoscopy and histology remain the golden standards for clinical detection and diagnosis, respectively. However, the intermediate stage of gross pathology hinders the digital flow. Gross pathology is a visual assessment stage that includes the manual preparation of ex-vivo tissue for histology. Despite advances in medical technology, gross pathology lacks digitization, automation, and standardization. At the same time, it is time-consuming and dependent on the skill of the pathologist. However, pigmented skin lesions (PSL), a group of pathologies with a high incidence rate, are rich in color information. Hyperspectral imaging (HSI) is a noninvasive imaging technique that can capture spectral signatures of a tissue sample, capable of discriminating malignancies.3 A noninvasive, nonionizing optical biopsy based on HSI could speed up diagnosis, reduce the number of re-excisions, enable telepathology, and increase treatment effectiveness.
- 31 October 2022
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 27 (10) , 106007-22
Open AccessHyperspectral imaging for tumor segmentation on pigmented skin lesions
In Japan, incidence and mortality rates of skin cancer have increased in the past decades, despite adjustments for an aging population.1 At the same time, the American Cancer Society estimates that malignant skin tumors comprise the majority of all cancers in the United States.2 Early diagnosis is key for the effective treatment of skin tumors. Dermoscopy and histology remain the golden standards for clinical detection and diagnosis, respectively. However, the intermediate stage of gross pathology hinders the digital flow. Gross pathology is a visual assessment stage that includes the manual preparation of ex-vivo tissue for histology. Despite advances in medical technology, gross pathology lacks digitization, automation, and standardization. At the same time, it is time-consuming and dependent on the skill of the pathologist. However, pigmented skin lesions (PSL), a group of pathologies with a high incidence rate, are rich in color information. Hyperspectral imaging (HSI) is a noninvasive imaging technique that can capture spectral signatures of a tissue sample, capable of discriminating malignancies.3 A noninvasive, nonionizing optical biopsy based on HSI could speed up diagnosis, reduce the number of re-excisions, enable telepathology, and increase treatment effectiveness.
- Open Access
- 8 June 2022
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 27 (06) , 060901-28
Hyperspectral and multispectral image processing for gross-level tumor detection in skin lesions: a systematic review
More people are diagnosed with skin cancer each year in the United States than all other cancers combined.1 Malignant melanoma (MM) and other skin cancer crude mortality rates increased up to twofold in Japan from 1999 to 2014, independent of the aging of the Japanese population.2 The increased workload demands increased effectiveness in the diagnosis and management of malignant lesions. Skin lesions of concern are evaluated clinically and/or histopathologically. Gross pathology, the intermediate stage between the two, remains a manual, labor-intensive, and non-standardized step in the diagnostic work flow,3 despite recent advances in digital pathology.4 Hyperspectral imaging (HSI) is an imaging modality that can assist gross pathology. HSI captures spectral signatures of tissue pixels dependent on substance concentrations inside the tissue and has been proven useful for staining and standardization in histopathology.5 In the past decade, several research groups have investigated applications of HSI and multispectral imaging (MSI) at the gross pathology stage, toward diagnosis and tumor classification of skin lesions, aiming to achieve noninvasive optical biopsies and increase diagnosis speed.
- 8 June 2022
- Published by SPIE-Intl Soc Optical Eng in Journal of Biomedical Optics
- Vol. 27 (06) , 060901-28
Open AccessHyperspectral and multispectral image processing for gross-level tumor detection in skin lesions: a systematic review
More people are diagnosed with skin cancer each year in the United States than all other cancers combined.1 Malignant melanoma (MM) and other skin cancer crude mortality rates increased up to twofold in Japan from 1999 to 2014, independent of the aging of the Japanese population.2 The increased workload demands increased effectiveness in the diagnosis and management of malignant lesions. Skin lesions of concern are evaluated clinically and/or histopathologically. Gross pathology, the intermediate stage between the two, remains a manual, labor-intensive, and non-standardized step in the diagnostic work flow,3 despite recent advances in digital pathology.4 Hyperspectral imaging (HSI) is an imaging modality that can assist gross pathology. HSI captures spectral signatures of tissue pixels dependent on substance concentrations inside the tissue and has been proven useful for staining and standardization in histopathology.5 In the past decade, several research groups have investigated applications of HSI and multispectral imaging (MSI) at the gross pathology stage, toward diagnosis and tumor classification of skin lesions, aiming to achieve noninvasive optical biopsies and increase diagnosis speed.