A5004690585- Cardiac Valve Diseases and Treatments
- Generative Adversarial Networks and Image Synthesis
- Welding Techniques and Residual Stresses
- Adversarial Robustness in Machine Learning
- Industrial Vision Systems and Defect Detection
- Statistical Methods and Inference
- Radiomics and Machine Learning in Medical Imaging
- Advanced Steganography and Watermarking Techniques
- Cardiac Imaging and Diagnostics
- Digital Media Forensic Detection
- Advanced Causal Inference Techniques
- Medical Image Segmentation Techniques
- Sparse and Compressive Sensing Techniques
- Geological Modeling and Analysis
- Stellar, planetary, and galactic studies
- Advanced Image and Video Retrieval Techniques
- Galaxies: Formation, Evolution, Phenomena
- Advanced MRI Techniques and Applications
- Astronomy and Astrophysical Research
- Image and Signal Denoising Methods
- Natural Language Processing Techniques
- Domain Adaptation and Few-Shot Learning
- Explainable Artificial Intelligence (XAI)
- Lung Cancer Diagnosis and Treatment
- Medical Imaging and Analysis
Hitachi (Japan)
2017-2023
Line Corporation (Japan)
2023
Waseda University
2015
We report the spectroscopic confirmation of a merging pair massive quiescent galaxies at z=3.44. Using JWST observations, we confirm that two lie projected separation 4.5 kpc with velocity offset ∼ 680 (δ_z 0.01). The resides in core known rich overdensity galaxies, dubbed "Cosmic Vine". For both members, modeling spectral energy distributions and faint rest-frame optical emission lines indicate high stellar masses (M_⋆/M_⊙) ∼10.9) suppressed star formation (log <-10), more than an order...
Image manipulation techniques have drawn growing concerns as manipulated images might cause morality and security problems. Various methods been proposed to detect manipulations achieved promising performance. However, these be vulnerable adversarial attacks. In this work, we design an Adversarial Manipulation Generation (AMG) task explore the vulnerability of image detectors. We first propose optimal loss function extend existing attacks generate examples. observe that spatial large...
The extraction of six standard planes in 3-D cardiac ultrasound plays an important role clinical examination to analyze function. A guideline-based learning method for efficient and accurate plane is proposed. guideline determines appropriate operation steps examinations. idea incorporating machine approaches into each stage the guideline. First, Hough forest with hierarchical search applied feature point detection. Second, initial are determined using anatomical regularities according...
This paper presents our 3rd place solution in both Descriptor Track and Matching of the Meta AI Video Similarity Challenge (VSC2022), a competition aimed at detecting video copies. Our approach builds upon existing image copy detection techniques incorporates several strategies to exploit on properties data, resulting simple yet powerful solution. By employing proposed method, we achieved substantial improvements accuracy compared baseline results (Descriptor Track: 38% improvement, 60%...
Diffusion Models (DMs) have shown remarkable capabilities in various image-generation tasks. However, there are growing concerns that DMs could be used to imitate unauthorized creations and thus raise copyright issues. To address this issue, we propose a novel framework embeds personal watermarks the generation of adversarial examples. Such examples can force generate images with visible prevent from imitating images. We construct generator based on conditional networks design three losses...
Dynamic MRI is widely used for many clinical exams but slow data acquisition becomes a serious problem. The application of Compressed Sensing (CS) demonstrated great potential to increase imaging speed. However, the performance CS largely depending on sparsity image sequence in transform domain, where there are still lot be improved. In this work, exploited by proposed Sparse Decomposition Learning (SDL) algorithm, which combination low-rank plus and Blind (BCS). With decomposition, only...
Abstract Estimating individual treatment effects (ITE) from observational data have become an important topic in various fields. In healthcare, determining the optimal time and amount of is crucial for improving each patient’s quality life. However, existing methods mainly focused on estimating effect under static setting with categorical treatments. this work, we proposed Recurrent Continuous Adversarial Balancing (RCAB), a method that incorporates adversarial learning into...
Large web crawl datasets have already played an important role in learning multimodal features with high generalization capabilities. However, there are still very limited studies investigating the details or improvements of data design. Recently, a DataComp challenge has been designed to propose best training fixed models. This paper presents our solution both filtering track and BYOD challenge. Our adopts large models CLIP BLIP-2 filter modify data, utilize external along bag tricks...
In recent years, document processing has flourished and brought numerous benefits. However, there been a significant rise in reported cases of forged images. Specifically, advancements deep neural network (DNN) methods for generative tasks may amplify the threat forgery. Traditional approaches images created by prevalent copy-move are unsuitable against those DNN-based methods, as we have verified. To address this issue, construct training dataset forgery images, named FD-VIED, emulating...
High-speed acquisition of ultrasound volume data is needed for fetal cardiac diagnosis. The heartbeat healthy fetus 120-160 times per second. Therefore, the acquiring method based on plane wave compounding has been developed to achieve both rate and image quality. However, in conventional method, sufficient quality couldn't be obtained when speed about 150 volumes Compressed Sensing applied improve reduce number previous work [1], but huge memory (458GB) required. In this paper, we propose a...