-
Event Classification with Multi-step Machine Learning
Authors:
Masahiko Saito,
Tomoe Kishimoto,
Yuya Kaneta,
Taichi Itoh,
Yoshiaki Umeda,
Junichi Tanaka,
Yutaro Iiyama,
Ryu Sawada,
Koji Terashi
Abstract:
The usefulness and value of Multi-step Machine Learning (ML), where a task is organized into connected sub-tasks with known intermediate inference goals, as opposed to a single large model learned end-to-end without intermediate sub-tasks, is presented. Pre-optimized ML models are connected and better performance is obtained by re-optimizing the connected one. The selection of an ML model from sev…
▽ More
The usefulness and value of Multi-step Machine Learning (ML), where a task is organized into connected sub-tasks with known intermediate inference goals, as opposed to a single large model learned end-to-end without intermediate sub-tasks, is presented. Pre-optimized ML models are connected and better performance is obtained by re-optimizing the connected one. The selection of an ML model from several small ML model candidates for each sub-task has been performed by using the idea based on Neural Architecture Search (NAS). In this paper, Differentiable Architecture Search (DARTS) and Single Path One-Shot NAS (SPOS-NAS) are tested, where the construction of loss functions is improved to keep all ML models smoothly learning. Using DARTS and SPOS-NAS as an optimization and selection as well as the connections for multi-step machine learning systems, we find that (1) such a system can quickly and successfully select highly performant model combinations, and (2) the selected models are consistent with baseline algorithms, such as grid search, and their outputs are well controlled.
△ Less
Submitted 4 June, 2021;
originally announced June 2021.
-
An Improvement of Object Detection Performance using Multi-step Machine Learnings
Authors:
Tomoe Kishimoto,
Masahiko Saito,
Junichi Tanaka,
Yutaro Iiyama,
Ryu Sawada,
Koji Terashi
Abstract:
Connecting multiple machine learning models into a pipeline is effective for handling complex problems. By breaking down the problem into steps, each tackled by a specific component model of the pipeline, the overall solution can be made accurate and explainable. This paper describes an enhancement of object detection based on this multi-step concept, where a post-processing step called the calibr…
▽ More
Connecting multiple machine learning models into a pipeline is effective for handling complex problems. By breaking down the problem into steps, each tackled by a specific component model of the pipeline, the overall solution can be made accurate and explainable. This paper describes an enhancement of object detection based on this multi-step concept, where a post-processing step called the calibration model is introduced. The calibration model consists of a convolutional neural network, and utilizes rich contextual information based on the domain knowledge of the input. Improvements of object detection performance by 0.8-1.9 in average precision metric over existing object detectors have been observed using the new model.
△ Less
Submitted 19 January, 2021;
originally announced January 2021.
-
Toward Co-existing Database Schemas based on Bidirectional Transformation
Authors:
Jumpei Tanaka,
Van-Dang Tran,
Hiroyuki kato,
Zhenjiang Hu
Abstract:
According to strong demands for rapid and reliable software delivery, co-existing database schema versions with multiple application versions are reality to contribute them. Current database management systems do not support co-existing schema versions in one database. Although a design of co-existing schema based on updatable view tables was previously proposed, its flexibility is limited due to…
▽ More
According to strong demands for rapid and reliable software delivery, co-existing database schema versions with multiple application versions are reality to contribute them. Current database management systems do not support co-existing schema versions in one database. Although a design of co-existing schema based on updatable view tables was previously proposed, its flexibility is limited due to pre-defined several restrictions to achieve data synchronization among schemas and handling independent unsynchronized data in each schema. In this preliminary report, we present a new approach for co-existing schemas based on bidirectional transformation. We explain the required properties to realize co-existing schemas, bidirectionality and totality. We show that the co-existing schemas can be implemented systematically by applying putback-based bidirectional transformation to satisfy both the bidirectionality and the totality. While the bidirectionality can be satisfied by applying bidirectional transformation, to satisfy the totality, extra functions need to be introduced. How to derive these extra functions is presented.
△ Less
Submitted 30 October, 2019; v1 submitted 24 October, 2019;
originally announced October 2019.
-
Aiding autobiographic memory by using wearable devices
Authors:
Jingyi Wang,
Jiro Tanaka
Abstract:
In this paper, we investigate the effectiveness of two distinct techniques (Special Moment Approach & Spatial Frequency Approach) for reviewing the lifelogs, which were collected by lifeloggers who were willing to use a wearable camera and a bracelet simultaneously for two days. Generally, Special moment approach is a technique for extracting episodic events and Spatial frequency approach is a tec…
▽ More
In this paper, we investigate the effectiveness of two distinct techniques (Special Moment Approach & Spatial Frequency Approach) for reviewing the lifelogs, which were collected by lifeloggers who were willing to use a wearable camera and a bracelet simultaneously for two days. Generally, Special moment approach is a technique for extracting episodic events and Spatial frequency approach is a technique for associating visual with temporal and location information, especially heat map is applied as the spatial data for expressing frequency awareness. Based on that, the participants were asked to fill in two post-study questionnaires for evaluating the effectiveness of those two techniques and their combination. The preliminary result showed the positive potential of exploring individual lifelogs using our approaches.
△ Less
Submitted 5 June, 2017;
originally announced June 2017.
-
Interaction With Tilting Gestures In Ubiquitous Environments
Authors:
Ayman Atia,
Jiro Tanaka
Abstract:
In this paper, we introduce a tilting interface that controls direction based applications in ubiquitous environments. A tilt interface is useful for situations that require remote and quick interactions or that are executed in public spaces. We explored the proposed tilting interface with different application types and classified the tilting interaction techniques. Augmenting objects with sensor…
▽ More
In this paper, we introduce a tilting interface that controls direction based applications in ubiquitous environments. A tilt interface is useful for situations that require remote and quick interactions or that are executed in public spaces. We explored the proposed tilting interface with different application types and classified the tilting interaction techniques. Augmenting objects with sensors can potentially address the problem of the lack of intuitive and natural input devices in ubiquitous environments. We have conducted an experiment to test the usability of the proposed tilting interface to compare it with conventional input devices and hand gestures. The experiment results showed greater improvement of the tilt gestures in comparison with hand gestures in terms of speed, accuracy, and user satisfaction.
△ Less
Submitted 29 July, 2010;
originally announced July 2010.
-
Logic with Verbs
Authors:
Jun Tanaka
Abstract:
The aim of this paper is to introduce a logic in which nouns and verbs are handled together as a deductive reasoning, and also to observe the relationship between nouns and verbs as well as between logics and conversations.
The aim of this paper is to introduce a logic in which nouns and verbs are handled together as a deductive reasoning, and also to observe the relationship between nouns and verbs as well as between logics and conversations.
△ Less
Submitted 14 December, 2009; v1 submitted 12 September, 2009;
originally announced September 2009.
-
Modern Set
Authors:
Jun Tanaka
Abstract:
In this paper, we intend to generalize the classical set theory as much as possible. we will do this by freeing sets from the regular properties of classical sets; e.g., the law of excluded middle, the law of non-contradiction, the distributive law, the commutative law,etc....
In this paper, we intend to generalize the classical set theory as much as possible. we will do this by freeing sets from the regular properties of classical sets; e.g., the law of excluded middle, the law of non-contradiction, the distributive law, the commutative law,etc....
△ Less
Submitted 18 June, 2008;
originally announced June 2008.
-
A data Grid testbed environment in Gigabit WAN with HPSS
Authors:
Atsushi Manabe,
Kohki Ishikawa,
Yoshihiko Itoh,
Setsuya Kawabata,
Tetsuro Mashimo,
Youhei Morita,
Hiroshi Sakamoto,
Takashi Sasaki,
Hiroyuki Sato,
Junichi Tanaka,
Ikuo Ueda,
Yoshiyuki Watase,
Satomi Yamamoto,
Shigeo Yashiro
Abstract:
For data analysis of large-scale experiments such as LHC Atlas and other Japanese high energy and nuclear physics projects, we have constructed a Grid test bed at ICEPP and KEK. These institutes are connected to national scientific gigabit network backbone called SuperSINET. In our test bed, we have installed NorduGrid middleware based on Globus, and connected 120TB HPSS at KEK as a large scale…
▽ More
For data analysis of large-scale experiments such as LHC Atlas and other Japanese high energy and nuclear physics projects, we have constructed a Grid test bed at ICEPP and KEK. These institutes are connected to national scientific gigabit network backbone called SuperSINET. In our test bed, we have installed NorduGrid middleware based on Globus, and connected 120TB HPSS at KEK as a large scale data store. Atlas simulation data at ICEPP has been transferred and accessed using SuperSINET. We have tested various performances and characteristics of HPSS through this high speed WAN. The measurement includes comparison between computing and storage resources are tightly coupled with low latency LAN and long distant WAN.
△ Less
Submitted 3 September, 2003; v1 submitted 12 June, 2003;
originally announced June 2003.
