-
Mixbiotic society measures: Comparison of organizational structures based on communication simulation
Authors:
Takeshi Kato,
Jyunichi Miyakoshi,
Tadayuki Matsumura,
Yasuyuki Kudo,
Ryuji Mine,
Hiroyuki Mizuno,
Yasuo Deguchi
Abstract:
The philosophical world has proposed the concept of "mixbiotic society," in which individuals with freedom and diverse values mix and mingle to recognize their respective "fundamental incapability" each other and sublimate into solidarity, toward solving the issues of social isolation and fragmentation. Based on this concept, the mixbiotic society measures have been proposed to evaluate dynamic co…
▽ More
The philosophical world has proposed the concept of "mixbiotic society," in which individuals with freedom and diverse values mix and mingle to recognize their respective "fundamental incapability" each other and sublimate into solidarity, toward solving the issues of social isolation and fragmentation. Based on this concept, the mixbiotic society measures have been proposed to evaluate dynamic communication patterns with reference to classification in cellular automata and particle reaction-diffusion that simulate living phenomena. In this paper, we applied these measures to five typologies of organizational structure (Red: impulsive, Amber: adaptive, Orange: achievement, Green: pluralistic, and Teal: evolutionary) and evaluated their features. Specifically, we formed star, tree, tree+jumpers, tree+more jumpers, and small-world type networks corresponding to each of five typologies, conducted communication simulations on these networks, and calculated values for mixbiotic society measures. The results showed that Teal organization has the highest value of the mixism measure among mixbiotic society measures, i.e., it balances similarity (mixing) and dissimilarity (mingling) in communication, and is living and mixbiotic between order and chaos. Measures other than mixism showed that in Teal organization, information is not concentrated in a central leader and that communication takes place among various members. This evaluation of organizational structures shows that the mixbiotic society measures is also useful for assessing organizational change. In the future, these measures will be used not only in business organizations, but also in digital democratic organizations and platform cooperatives in conjunction with information technology.
△ Less
Submitted 28 July, 2023;
originally announced July 2023.
-
Social Co-OS: Cyber-Human Social Co-Operating System
Authors:
Takeshi Kato,
Yasuyuki Kudo,
Junichi Miyakoshi,
Misa Owa,
Yasuhiro Asa,
Takashi Numata,
Ryuji Mine,
Hiroyuki Mizuno
Abstract:
The novel concept of a Cyber-Human Social System (CHSS) and a diverse and pluralistic 'mixed-life society' is proposed, wherein cyber and human societies commit to each other. This concept enhances the Cyber-Physical System (CPS), which is associated with the current Society 5.0, a social vision realised through the fusion of cyber (virtual) and physical (real) spaces following information society…
▽ More
The novel concept of a Cyber-Human Social System (CHSS) and a diverse and pluralistic 'mixed-life society' is proposed, wherein cyber and human societies commit to each other. This concept enhances the Cyber-Physical System (CPS), which is associated with the current Society 5.0, a social vision realised through the fusion of cyber (virtual) and physical (real) spaces following information society (Society 4.0 and Industry 4.0). Moreover, the CHSS enhances the Human-CPS, the Human-in-the-Loop CPS (HiLCPS), and the Cyber-Human System by intervening in individual behaviour pro-socially and supporting consensus building. As a form of architecture that embodies the CHSS concept, the Cyber-Human Social Co-Operating System (Social Co-OS) that combines cyber and human societies is shown. In this architecture, the cyber and human systems cooperate through the fast loop (operation and administration) and slow loop (consensus and politics). Furthermore, the technical content and current implementation of the basic functions of the Social Co-OS are described. These functions consist of individual behavioural diagnostics, interventions in the fast loop, group decision diagnostics and consensus building in the slow loop. Subsequently, this system will contribute to mutual aid communities and platform cooperatives.
△ Less
Submitted 21 September, 2022; v1 submitted 5 October, 2021;
originally announced October 2021.
-
Warp-Refine Propagation: Semi-Supervised Auto-labeling via Cycle-consistency
Authors:
Aditya Ganeshan,
Alexis Vallet,
Yasunori Kudo,
Shin-ichi Maeda,
Tommi Kerola,
Rares Ambrus,
Dennis Park,
Adrien Gaidon
Abstract:
Deep learning models for semantic segmentation rely on expensive, large-scale, manually annotated datasets. Labelling is a tedious process that can take hours per image. Automatically annotating video sequences by propagating sparsely labeled frames through time is a more scalable alternative. In this work, we propose a novel label propagation method, termed Warp-Refine Propagation, that combines…
▽ More
Deep learning models for semantic segmentation rely on expensive, large-scale, manually annotated datasets. Labelling is a tedious process that can take hours per image. Automatically annotating video sequences by propagating sparsely labeled frames through time is a more scalable alternative. In this work, we propose a novel label propagation method, termed Warp-Refine Propagation, that combines semantic cues with geometric cues to efficiently auto-label videos. Our method learns to refine geometrically-warped labels and infuse them with learned semantic priors in a semi-supervised setting by leveraging cycle consistency across time. We quantitatively show that our method improves label-propagation by a noteworthy margin of 13.1 mIoU on the ApolloScape dataset. Furthermore, by training with the auto-labelled frames, we achieve competitive results on three semantic-segmentation benchmarks, improving the state-of-the-art by a large margin of 1.8 and 3.61 mIoU on NYU-V2 and KITTI, while matching the current best results on Cityscapes.
△ Less
Submitted 27 September, 2021;
originally announced September 2021.
-
Hierarchical Lovász Embeddings for Proposal-free Panoptic Segmentation
Authors:
Tommi Kerola,
Jie Li,
Atsushi Kanehira,
Yasunori Kudo,
Alexis Vallet,
Adrien Gaidon
Abstract:
Panoptic segmentation brings together two separate tasks: instance and semantic segmentation. Although they are related, unifying them faces an apparent paradox: how to learn simultaneously instance-specific and category-specific (i.e. instance-agnostic) representations jointly. Hence, state-of-the-art panoptic segmentation methods use complex models with a distinct stream for each task. In contra…
▽ More
Panoptic segmentation brings together two separate tasks: instance and semantic segmentation. Although they are related, unifying them faces an apparent paradox: how to learn simultaneously instance-specific and category-specific (i.e. instance-agnostic) representations jointly. Hence, state-of-the-art panoptic segmentation methods use complex models with a distinct stream for each task. In contrast, we propose Hierarchical Lovász Embeddings, per pixel feature vectors that simultaneously encode instance- and category-level discriminative information. We use a hierarchical Lovász hinge loss to learn a low-dimensional embedding space structured into a unified semantic and instance hierarchy without requiring separate network branches or object proposals. Besides modeling instances precisely in a proposal-free manner, our Hierarchical Lovász Embeddings generalize to categories by using a simple Nearest-Class-Mean classifier, including for non-instance "stuff" classes where instance segmentation methods are not applicable. Our simple model achieves state-of-the-art results compared to existing proposal-free panoptic segmentation methods on Cityscapes, COCO, and Mapillary Vistas. Furthermore, our model demonstrates temporal stability between video frames.
△ Less
Submitted 8 June, 2021;
originally announced June 2021.
-
Unsupervised Adversarial Learning of 3D Human Pose from 2D Joint Locations
Authors:
Yasunori Kudo,
Keisuke Ogaki,
Yusuke Matsui,
Yuri Odagiri
Abstract:
The task of three-dimensional (3D) human pose estimation from a single image can be divided into two parts: (1) Two-dimensional (2D) human joint detection from the image and (2) estimating a 3D pose from the 2D joints. Herein, we focus on the second part, i.e., a 3D pose estimation from 2D joint locations. The problem with existing methods is that they require either (1) a 3D pose dataset or (2) 2…
▽ More
The task of three-dimensional (3D) human pose estimation from a single image can be divided into two parts: (1) Two-dimensional (2D) human joint detection from the image and (2) estimating a 3D pose from the 2D joints. Herein, we focus on the second part, i.e., a 3D pose estimation from 2D joint locations. The problem with existing methods is that they require either (1) a 3D pose dataset or (2) 2D joint locations in consecutive frames taken from a video sequence. We aim to solve these problems. For the first time, we propose a method that learns a 3D human pose without any 3D datasets. Our method can predict a 3D pose from 2D joint locations in a single image. Our system is based on the generative adversarial networks, and the networks are trained in an unsupervised manner. Our primary idea is that, if the network can predict a 3D human pose correctly, the 3D pose that is projected onto a 2D plane should not collapse even if it is rotated perpendicularly. We evaluated the performance of our method using Human3.6M and the MPII dataset and showed that our network can predict a 3D pose well even if the 3D dataset is not available during training.
△ Less
Submitted 22 March, 2018;
originally announced March 2018.
