Shoichiro Takeda

Shoichiro Takeda

Auto-generated profile, based on publicly available information

Shoichiro Takeda

NTT Human Informatics Laboratories, NTT Corporation, Yokosuka, Japan

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 August 2021 in IEEE Transactions on Circuits and Systems for Video Technology
Verified
Journal Article
Closed access
SSSatoshi Suzuki
Satoshi Suzuki
STShoichiro Takeda
Shoichiro Takeda
MTMotohiro Takagi
Motohiro Takagi
RTRyuichi Tanida
Ryuichi Tanida
HKHideaki Kimata
Hideaki Kimata
HSHayaru Shouno
Hayaru Shouno
Collaborative Intelligence is a new paradigm that splits a deep neural network (DNN) into an edge and cloud for deploying a DNN-based image recognition application. In this paradigm, deep features, which are the outputs of the edge DNN, are compressed and transmitted to the cloud DNN. Because the deep features have a number of responses that are similar to each other, for efficient compression, previous methods spatially arrange and compress the deep features as an image to utilize the similarity as a spatial correlation. However, if the deep features are arranged in not only spatial but also temporal directions like those in a video, it may be possible to compress them more efficiently by increasing a temporal correlation. To explore this possibility, we propose a “spatio-temporal arrangement”. This method spatially arranges the deep features as images and temporally arranges them as a video with a novel ordering search algorithm. Our method effectively increases the spatial and temporal correlations hidden in the deep features and achieves high compression efficiency compared with the previous methods. Experimental results demonstrate the compression efficiency of our method is better than that of the previous methods (1.50% to 4.98% on BD-Rate evaluation in a lossy setting). Our analysis shows that our method effectively increases the correlation when the input is an image with rich edges and textures.


24 June 2022 in IEEE Access
Verified
Journal Article
Open access
SSSatoshi Suzuki
Satoshi Suzuki
STShoichiro Takeda
Shoichiro Takeda
NMNaoki Makishima
Naoki Makishima
AAAtsushi Ando
Atsushi Ando
RMRyo Masumura
Ryo Masumura
HSHayaru Shouno
Hayaru Shouno
Anti-aliased convolutional neural networks (CNNs) are models that introduce blur filters to intermediate representations of CNNs to achieve high accuracy in image recognition tasks. A promising way to prepare a new anti-aliased CNN is to introduce blur filters to the intermediate representations of pre-trained (non anti-aliased) CNNs, since many researchers have released them online. Although this scheme can build the new anti-aliased CNN easily, the blur filters drastically degrade the pre-trained representations. Therefore, to take full advantage of the benefits of introducing blur filters, fine-tuning using massive amounts of training data is often required. This can be problematic because the training data is often limited. In such a “data-limited” situation, the fine-tuning does not bring about a high performance because it induces overfitting to the limited training data. To tackle this problem, we propose “knowledge transferred fine-tuning.” Knowledge transfer is a technique that utilizes the representations of a pre-trained model to help ensure generalization in data-limited situations. Inspired by this concept, we transfer knowledge from intermediate representations in a pre-trained CNN to an anti-aliased CNN while fine-tuning. The key idea of our method is to transfer only the essential knowledge for image recognition in the pre-trained CNN using two types of loss: pixel-level loss and global-level loss. The former loss transfers the detailed knowledge from the pre-trained CNN, but this knowledge may contain “aliased” non-essential knowledge. The latter loss, on the other hand, is designed to increase when the pixel-level loss transfers non-essential knowledge while ignoring the essential knowledge, i.e., it penalizes the pixel-level loss. Experimental results demonstrate that the proposed method using just 25 training images per class on ImageNet 2012 can achieve higher accuracy than a conventional pre-trained CNN.

8 August 2023 in Neural Networks
Verified
Journal Article
Open access
SSSatoshi Suzuki
Satoshi Suzuki
STShoichiro Takeda
Shoichiro Takeda
RTRyuichi Tanida
Ryuichi Tanida
YBYukihiro Bandoh
Yukihiro Bandoh
HSHayaru Shouno
Hayaru Shouno
In image classification, a deep neural network (DNN) that is trained on undistorted images constitutes an effective decision boundary. Unfortunately, this boundary does not support distorted images, such as noisy or blurry ones, leading to accuracy drop-off. As a simple approach for classifying distorted images as well as undistorted ones, previous methods have optimized the trained DNN again on both kinds of images. However, in these methods, the decision boundary may become overly complicated during optimization because there is no regularization of the decision boundary. Consequently, this decision boundary limits efficient optimization. In this paper, we study a simple yet effective decision boundary for distorted image classification through the use of a novel loss, called a "neural activation pattern matching (NAPM) loss". The NAPM loss is based on recent findings that the decision boundary is a piecewise linear function, where each linear segment is constructed from a neural activation pattern in the DNN when an image is fed to it. The NAPM loss extracts the neural activation patterns when the distorted image and its undistorted version are fed to the DNN and then matches them with each other via the sigmoid cross-entropy. Therefore, it constrains the DNN to classify the distorted image and its undistorted version by the same linear segment. As a result, our loss accelerates efficient optimization by preventing the decision boundary from becoming overly complicated. Our experiments demonstrate that our loss increases the accuracy of the previous methods in all conditions evaluated.

4 July 2024 in IEEE Access
Verified
Journal Article
Open access
TNTakasuke Nagai
Takasuke Nagai
STShoichiro Takeda
Shoichiro Takeda
SSSatoshi Suzuki
Satoshi Suzuki
HSHitoshi Seshimo
Hitoshi Seshimo
Action quality assessment (AQA) is a task for assessing a specific action quality in videos. Since existing AQA datasets provide only two-dimensional (2D) video data captured from fewer viewpoints, existing AQA methods based on deep neural networks (DNNs) struggle to assess complex three-dimensional (3D) actions accurately, and their robustness against diversified viewpoints remains unknown. We created a dataset called multimodal in-the-wild (MMW)-AQA in freestyle windsurfing that addresses these concerns. In addition to video data, MMW-AQA provides inertial measurement unit (IMU) and global positioning system (GPS) data. The 3D information of IMU data helps DNNs accurately assess complex 3D actions. Moreover, MMW-AQA provides wild video data captured by an unmanned aerial vehicle (UAV). These wild video data enable us to evaluate whether AQA methods can work well on diversified viewpoints. Furthermore, we also present the baseline multimodalization framework with a transformer-based fusion module. These frameworks multimodalize existing unimodal DNN models easily to assess action quality using multimodal data. Our experimental results demonstrate that multimodal data improves the AQA accuracy compared with unimodal video data.

24 February 2023 in IEEE Transactions on Signal and Information Processing over Networks
Verified
Journal Article
Closed access
KNKenta Niwa
Kenta Niwa
NUNaonori Ueda
Naonori Ueda
HSHiroshi Sawada
Hiroshi Sawada
AFAkinori Fujino
Akinori Fujino
STShoichiro Takeda
Shoichiro Takeda
WKW. Bastiaan Kleijn
W. Bastiaan Kleijn
A neural architecture to efficiently coordinate the transitions of state variables (states) over a graph is proposed. We consider the coordination of the time evolution of the state variables associated with the nodes on a graph. The states are associated with physical attributes of agents, e.g., the speed and/or location of vehicles. Efficient coordination then corresponds to the optimization of dynamics (how fast do vehicles travel) that may be subject to constraints (travel must be collision-free). The aim of this paper is (i) the formulation of learnable constrained dynamics which governs the transition of states under constraints over a graph and (ii) to show its industrial application. Firstly, we formulate continuous ordinary differential equations (ODEs), namely forward propagation of state transitions over a graph and backward propagation to optimize model parameters for the learnable constrained dynamics. Discretization of these continuous ODEs results in the neural architecture CoordiNet . Secondly, as an application of CoordiNet, we address a traffic coordination problem by learning constrained dynamics such that vehicles can travel as fast as possible without collisions. Simulation experiments of traffic coordination confirm that our method maximizes the vehicles' speed states while avoiding collisions.