レッドハット 朝活セミナー(1/15, 2/18)の下記セッションでの発表予定資料です。
「Red Hat Enterprise Linux OpenStack Platform環境でのDocker活用テクニック」
https://www.redhat.com/ja/about/events/red-hat-asakatsu-seminar-2016
レッドハット 朝活セミナー(1/15, 2/18)の下記セッションでの発表予定資料です。
「Red Hat Enterprise Linux OpenStack Platform環境でのDocker活用テクニック」
https://www.redhat.com/ja/about/events/red-hat-asakatsu-seminar-2016
Geeko Magazine: A Technical Magazine on openSUSE, edited on openSUSE―openSUSE...Fuminobu Takeyama
This is the same talk as openSUSE Conference 2019.
Japan openSUSE User Group publishes a technical magazine every half year. The title of the magazine is Geeko Magazine. It consists of technical articles on openSUSE and applications running on openSUSE. For example, the latest issue of the magazine contains articles like "Launching Kubernetes Cluster with Kubic in 10 minutes", "Accessing to Google Drive from openSUSE", and "How to enable HTTPS with Let's Encrypt." Since 2014, we have published 9 issues of Geeko Magazine.
In this talk, after explaining the culture of self-publishing in Japan, I will talk about the process from calling articles from the user group until distributing Geeko Magazine.
Another topic is our challenge: editing the magazine on openSUSE. Thereby, we cannot use popular desktop publishing (DTP) applications like Adobe InDesign. Instead of such applications, we have been using Scribus, an OSS DTP application to edit Geeko Magazine. It supports CMYK color and DTP data such as trim marks and bleed areas, required by print shops. However, in 2014, Scribus was not adequate for writing a Japanese document. This is because typesetting rules are much different from English etc. To publish Geeko Magazine, we went OSS way; we have improved Scribus one by one at every issue of the magazine in cooperation with the upstream community. I will talk about a brief summary of those problems we have resolved.
Geeko Magazine: A Technical Magazine on openSUSE, editied on openSUSEFuminobu Takeyama
Japan openSUSE User Group publishes a technical magazine every half year. The title of the magazine is Geeko Magazine. It consists of technical articles on openSUSE and applications running on openSUSE. For example, the latest issue of the magazine contains articles like "Launching Kubernetes Cluster with Kubic in 10 minutes", "Accessing to Google Drive from openSUSE", and "How to enable HTTPS with Let's Encrypt." Since 2014, we have published 9 issues of Geeko Magazine.
In this talk, after explaining the culture of self-publishing in Japan, I will talk about the process from calling articles from the user group until distributing Geeko Magazine.
Another topic is our challenge: editing the magazine on openSUSE. Thereby, we cannot use popular desktop publishing (DTP) applications like Adobe InDesign. Instead of such applications, we have been using Scribus, an OSS DTP application to edit Geeko Magazine. It supports CMYK color and DTP data such as trim marks and bleed areas, required by print shops. However, in 2014, Scribus was not adequate for writing a Japanese document. This is because typesetting rules are much different from English etc. To publish Geeko Magazine, we went OSS way; we have improved Scribus one by one at every issue of the magazine in cooperation with the upstream community. I will talk about a brief summary of those problems we have resolved.
What is necessary for the next input method framework?Fuminobu Takeyama
Input method framework is a software that has been used to input complex characters (e.g., Chinese characters, Hiragana, Hangul). Strictly speaking, the role of input method framework is bridging desktop applications and input method engines, which translate typed keys into complex characters. During this two decades, several input method frameworks including Kinput2, SCIM, UIM, Gcin, Fcitx, and IBus have been developed.
The situation surrounding text input method is changing. Firstly, new approaches such as software keyboard for touch screen and speech to text input are available on platforms other than Linux desktop.
Another change is the integration of input method into desktop application platforms. For example, IBus is now a part of GNOME desktop environment; Not only GNOME, Qt also include IBus support. Furthermore, Flatpak also uses a subset of IBus D-Bus interface for applications in Flatpak sandbox to communicate with an input method running on its host desktop.
Some people might think IBus is the defacto standard input method framework. However, quite many people prefer Fcitx or else due to the design issues of IBus. The latter people might think they are losing their freedom to select input method framework.
Now the speaker thinks that it is time to discuss the variety of input method frameworks is really necessary for the future Linux desktop environment. An input method framework itself does not provide much experience for users because its primary role is usually invisible to users. Thereby, how about bringing the war of input method framework and using our effort to improve input method engine, implementing the new approaches mentioned above, and supporting newer application platforms like Wayland?
In this talk, we would like to discuss what is necessary for the future input method framework by reviewing the design issues of IBus. The topics will be as the followings:
- Thin and high-level protocol with libraries avoiding code duplication
- The separation of responsibility between IBus daemon and plugins
- Importance of opened community
Note that the speaker is neither a developer of IBus nor GNOME. The attendees from GNOME community are welcome to improve this discussion.
セル生産方式におけるロボットの活用には様々な問題があるが,その一つとして 3 体以上の物体の組み立てが挙げられる.一般に,複数物体を同時に組み立てる際は,対象の部品をそれぞれロボットアームまたは治具でそれぞれ独立に保持することで組み立てを遂行すると考えられる.ただし,この方法ではロボットアームや治具を部品数と同じ数だけ必要とし,部品数が多いほどコスト面や設置スペースの関係で無駄が多くなる.この課題に対して音𣷓らは組み立て対象物に働く接触力等の解析により,治具等で固定されていない対象物が組み立て作業中に運動しにくい状態となる条件を求めた.すなわち,環境中の非把持対象物のロバスト性を考慮して,組み立て作業条件を検討している.本研究ではこの方策に基づいて,複数物体の組み立て作業を単腕マニピュレータで実行することを目的とする.このとき,対象物のロバスト性を考慮することで,仮組状態の複数物体を同時に扱う手法を提案する.作業対象としてパイプジョイントの組み立てを挙げ,簡易な道具を用いることで単腕マニピュレータで複数物体を同時に把持できることを示す.さらに,作業成功率の向上のために RGB-D カメラを用いた物体の位置検出に基づくロボット制御及び動作計画を実装する.
This paper discusses assembly operations using a single manipulator and a parallel gripper to simultaneously
grasp multiple objects and hold the group of temporarily assembled objects. Multiple robots and jigs generally operate
assembly tasks by constraining the target objects mechanically or geometrically to prevent them from moving. It is
necessary to analyze the physical interaction between the objects for such constraints to achieve the tasks with a single
gripper. In this paper, we focus on assembling pipe joints as an example and discuss constraining the motion of the
objects. Our demonstration shows that a simple tool can facilitate holding multiple objects with a single gripper.
【DLゼミ】XFeat: Accelerated Features for Lightweight Image Matchingharmonylab
公開URL:https://arxiv.org/pdf/2404.19174
出典:Guilherme Potje, Felipe Cadar, Andre Araujo, Renato Martins, Erickson R. ascimento: XFeat: Accelerated Features for Lightweight Image Matching, Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023)
概要:リソース効率に優れた特徴点マッチングのための軽量なアーキテクチャ「XFeat(Accelerated Features)」を提案します。手法は、局所的な特徴点の検出、抽出、マッチングのための畳み込みニューラルネットワークの基本的な設計を再検討します。特に、リソースが限られたデバイス向けに迅速かつ堅牢なアルゴリズムが必要とされるため、解像度を可能な限り高く保ちながら、ネットワークのチャネル数を制限します。さらに、スパース下でのマッチングを選択できる設計となっており、ナビゲーションやARなどのアプリケーションに適しています。XFeatは、高速かつ同等以上の精度を実現し、一般的なラップトップのCPU上でリアルタイムで動作します。