A new artificial life body: biologically inspired dynamic bipedal humanoid robots

Recently, a biologically inspired, bipedal, dynamic, humanoid robot was developed at the Artificial Life and Robotics Laboratory of Oita University. This bipedal humanoid robot is able to walk dynamically and to go up and down stairs. The central pattern generator developed produces various types of walking pattern. This robot has a pair of small CMOS color CCD cameras, a speaker, and a microphone in the head part, and will have a GPS, a portable telephone, and other sensors in the body part, so that the integration of locomotion and behavior to achieve specific demonstrations will be realized. This project develops dynamic mobility and the ability for autonomous recognition and navigation using the biological central nervous system, the brain system, and the real-time control system. Also, the design principles that demonstrate the dynamic interaction between neural and mechanical controls will be clarified. In Phase I, the platform of a small, bipedal, humanoid robot is used to develop autonomous locomotion and autonomous sensing and navigation. In Phase II of the project, an iteration on the platform design for human-size, bipedal, humanoid robots will be performed for operational testing. The development of bipedal humanoid robots that capture biological systems with unique principles and practices could dramatically increase their performance in tasks for national security needs.This work was presented in part at the 8th International Symposium on Artificial Life and Robotics, Oita, Japan, January 24–26, 2003     

A new approach to vision-based unsupervised learning of unexplored indoor environment for autonomous land vehicle navigation

A vision-based approach to unsupervised learning of the indoor environment for autonomous land vehicle (ALV) navigation is proposed. The ALV may, without human's involvement, self-navigate systematically in an unexplored closed environment, collect the information of the environment features, and then build a top-view map of the environment for later planned navigation or other applications. The learning system consists of three subsystems: a feature location subsystem, a model management subsystem, and an environment exploration subsystem. The feature location subsystem processes input images, and calculates the locations of the local features and the ALV by model matching techniques. To facilitate feature collection, two laser markers are mounted on the vehicle which project laser light on the corridor walls to form easily detectable line and corner features. The model management subsystem attaches the local model into a global one by merging matched corner pairs as well as line segment pairs. The environment exploration subsystem guides the ALV to explore the entire navigation environment by using the information of the learned model and the current ALV location. The guidance scheme is based on the use of a pushdown transducer derived from automata theory. A prototype learning system was implemented on a real vehicle, and simulations and experimental results in real environments show the feasibility of the proposed approach.     

Living in a partially structured environment: How to bypass the limitations of classical reinforcement techniques

In this paper, we propose an unsupervised neural network allowing a robot to learn sensory-motor associations with a delayed reward. The robot task is to learn the “meaning” of pictograms in order to “survive” in a maze. First, we introduce a new neural conditioning rule probabilistic conditioning rule (PCR) allowing us to test hypotheses (associations between visual categories and movements) during a given time span. Second, we describe a real maze experiment with our mobile robot. We propose a neural architecture overcoming the difficulty to build visual categories dynamically while associating them to movements. Third, we propose to use our algorithm on a simulation in order to test it exhaustively. We give the results for different kinds of mazes and we compare our system to an adapted version of the Q-learning algorithm. Finally, we conclude by showing the limitations of approaches that do not take into account the intrinsic complexity of a reasoning based on image recognition.     

Neocognitron: A new algorithm for pattern recognition tolerant of deformations and shifts in position

Suggested by the structure of the visual nervous system, a new algorithm is proposed for pattern recognition. This algorithm can be realized with a multilayered network consisting of neuron-like cells. The network, “neocognitron”, is self-organized by unsupervised learning, and acquires the ability to recognize stimulus patterns according to the differences in their shapes: Any patterns which we human beings judge to be alike are also judged to be of the same category by the neocognitron. The neocognitron recognizes stimulus patterns correctly without being affected by shifts in position or even by considerable distortions in shape of the stimulus patterns.     

深度域适应综述: 一般情况与复杂情况

信息时代产生的大量数据使机器学习技术成功地应用于许多领域. 大多数机器学习技术需要满足训练集与测试集独立同分布的假设, 但在实际应用中这个假设很难满足. 域适应是一种在训练集和测试集不满足独立同分布条件下的机器学习技术. 一般情况下的域适应只适用于源域目标域特征空间与标签空间都相同的情况, 然而实际上这个条件很难满足. 为了增强域适应技术的适用性, 复杂情况下的域适应逐渐成为研究热点, 其中标签空间不一致和复杂目标域情况下的域适应技术是近年来的新兴方向. 随着深度学习技术的崛起, 深度域适应已经成为域适应研究领域中的主流方法. 本文对一般情况与复杂情况下的深度域适应的研究进展进行综述, 对其缺点进行总结, 并对其未来的发展趋势进行预测. 首先对迁移学习相关概念进行介绍, 然后分别对一般情况与复杂情况下的域适应、域适应技术的应用以及域适应方法性能的实验结果进行综述, 最后对域适应领域的未来发展趋势进行展望并对全文内容进行总结.     

Self-calibration based on invariant view recognition: Dynamic approach to navigation

We extend a dynamic approach of behavior generation to the representation of spatial information. Two levels of dynamics integrate dead-reckoning, dominant far from home bases, and piloting, dominant near home bases. When the view-based piloting system recognizes a home base, visual place information recalibrates the dead-reckoning system, inverting the hierarchical ordering of the two dynamic levels by time scale inversion. Reference views taken at discrete home bases are recognized invariantly under rotation of views. This process yields compass information. Continuous translational information is obtained as a neural place representation built from view correlations with a scattered set of local views. This self-calibrating cognitive map couples into a dynamics of heading direction integrating the behaviors of obstacle avoidance and target acquisition. Targets can be designated in terms of the cognitive map. We demonstrate the dynamical model in simulation.     

Neural Network Controller against Environment: A Coevolutive approach to Generalize Robot Navigation Behavior

In this paper, a new coevolutive method, called Uniform Coevolution, is introduced to learn weights of a neural network controller in autonomous robots. An evolutionary strategy is used to learn high-performance reactive behavior for navigation and collisions avoidance. The introduction of coevolutive over evolutionary strategies allows evolving the environment, to learn a general behavior able to solve the problem in different environments. Using a traditional evolutionary strategy method, without coevolution, the learning process obtains a specialized behavior. All the behaviors obtained, with/without coevolution have been tested in a set of environments and the capability of generalization is shown for each learned behavior. A simulator based on a mini-robot Khepera has been used to learn each behavior. The results show that Uniform Coevolution obtains better generalized solutions to examples-based problems.     

TurSOM: A paradigm bridging Turing's unorganized machines and self-organizing maps demonstrating dual self-organization

Self-organization is a widely used technique in unsupervised learning and data analysis, largely exemplified by k-means clustering, self-organizing maps (SOM) and adaptive resonance theory.In this paper we present a new algorithm: TurSOM, inspired by Turing's unorganized machines and Kohonen's SOM. Turing's unorganized machines are an early model of neural networks characterized by self-organizing connections, as opposed to self-organizing neurons in SOM.TurSOM introduces three new mechanisms to facilitate both neuron and connection self-organization. These mechanisms are: a connection learning rate, connection reorganization, and a neuron responsibility radius.TurSOM is implemented in a 1-dimensional network (i.e. chain of neurons) to exemplify the theoretical implications of these features. In this paper we demonstrate that TurSOM is superior to the classical SOM algorithm in several ways: (1) speed until convergence; (2) independent clusters; and (3) tangle-free networks.     

Archetypoids: A new approach to define representative archetypal data

The new concept archetypoids is introduced. Archetypoid analysis represents each observation in a dataset as a mixture of actual observations in the dataset, which are pure type or archetypoids. Unlike archetype analysis, archetypoids are real observations, not a mixture of observations. This is relevant when existing archetypal observations are needed, rather than fictitious ones. An algorithm is proposed to find them and some of their theoretical properties are introduced. It is also shown how they can be obtained when only dissimilarities between observations are known (features are unavailable). Archetypoid analysis is illustrated in two design problems and several examples, comparing them with the archetypes, the nearest observations to them and other unsupervised methods.     

A modular design of Bayesian networks using expert knowledge: Context-aware home service robot

Recently, demand for service robots increases, and, particularly, one for personal service robots, which requires robot intelligence, will be expected to increase more. Accordingly, studies on intelligent robots are spreading all over the world. In this situation, we attempt to realize context-awareness for home robot while previous robot research focused on image processing, control and low-level context recognition. This paper uses probabilistic modeling for service robots to provide users with high-level context-aware services required in home environment, and proposes a systematic modeling approach for modeling a number of Bayesian networks. The proposed approach supplements uncertain sensor input using Bayesian network modeling and enhances the efficiency in modeling and reasoning processes using modular design based on domain knowledge. We verify the proposed method is useful as measuring the performance of context-aware module and conducting subjective test.     

Data mining and knowledge discovery in materials science and engineering: A polymer nanocomposites case study

In this study, data mining and knowledge discovery techniques were employed to validate their efficacy in acquiring information about the viscoelastic properties of vapor-grown carbon nanofiber (VGCNF)/vinyl ester (VE) nanocomposites solely from data derived from a designed experimental study. Formulation and processing factors (VGCNF type, use of a dispersing agent, mixing method, and VGCNF weight fraction) and testing temperature were utilized as inputs and the storage modulus, loss modulus, and tan delta were selected as outputs. The data mining and knowledge discovery algorithms and techniques included self-organizing maps (SOMs) and clustering techniques. SOMs demonstrated that temperature had the most significant effect on the output responses followed by VGCNF weight fraction. SOMs also showed how to prepare different VGCNF/VE nanocomposites with the same storage and loss modulus responses. A clustering technique, i.e., fuzzy C-means algorithm, was also applied to discover certain patterns in nanocomposite behavior after using principal component analysis as a dimensionality reduction technique. Particularly, these techniques were able to separate the nanocomposite specimens into different clusters based on temperature and tan delta features as well as to place the neat VE specimens (i.e., specimens containing no VGCNFs) in separate clusters. Most importantly, the results from data mining are consistent with previous response surface characterizations of this nanocomposite system. This work highlights the significance and utility of data mining and knowledge discovery techniques in the context of materials informatics.     

Spontaneous facial expression recognition: A robust metric learning approach

Spontaneous facial expression recognition is significantly more challenging than recognizing posed ones. We focus on two issues that are still under-addressed in this area. First, due to the inherent subtlety, the geometric and appearance features of spontaneous expressions tend to overlap with each other, making it hard for classifiers to find effective separation boundaries. Second, the training set usually contains dubious class labels which can hurt the recognition performance if no countermeasure is taken. In this paper, we propose a spontaneous expression recognition method based on robust metric learning with the aim of alleviating these two problems. In particular, to increase the discrimination of different facial expressions, we learn a new metric space in which spatially close data points have a higher probability of being in the same class. In addition, instead of using the noisy labels directly for metric learning, we define sensitivity and specificity to characterize the annotation reliability of each annotator. Then the distance metric and annotators' reliability is jointly estimated by maximizing the likelihood of the observed class labels. With the introduction of latent variables representing the true class labels, the distance metric and annotators' reliability can be iteratively solved under the Expectation Maximization framework. Comparative experiments show that our method achieves better recognition accuracy on spontaneous expression recognition, and the learned metric can be reliably transferred to recognize posed expressions.     

The more you learn,the less you store: Memory-controlled incremental SVM for visual place recognition

The capability to learn from experience is a key property for autonomous cognitive systems working in realistic settings. To this end, this paper presents an SVM-based algorithm, capable of learning model representations incrementally while keeping under control memory requirements. We combine an incremental extension of SVMs [43] with a method reducing the number of support vectors needed to build the decision function without any loss in performance [15] introducing a parameter which permits a user-set trade-off between performance and memory. The resulting algorithm is able to achieve the same recognition results as the original incremental method while reducing the memory growth. Our method is especially suited to work for autonomous systems in realistic settings. We present experiments on two common scenarios in this domain: adaptation in presence of dynamic changes and transfer of knowledge between two different autonomous agents, focusing in both cases on the problem of visual place recognition applied to mobile robot topological localization. Experiments in both scenarios clearly show the power of our approach.     

Motion planning for cooperative unicycle-type mobile robots with limited sensing ranges: A distributed receding horizon approach

This paper presents a decentralized motion planner for a team of nonholonomic mobile robots subject to constraints imposed by sensors and the communication network. The motion planning scheme consists of decentralized receding horizon planners that reside on each vehicle to achieve coordination among flocking agents. The advantage of the proposed algorithm is that each vehicle only requires local knowledge of its neighboring vehicles. The main requirement for designing an optimal conflict-free trajectory in a decentralized way is that each robot does not deviate too far from its presumed trajectory designed without taking the coupling constraints into account. A comparative study between the proposed algorithm and other existing algorithms is provided in order to show the advantages, especially in terms of computing time. Finally, experiments are performed on a team of three mobile robots to demonstrate the validity of the proposed approach.     

Detection and location of unsafe behaviour in digital images: A visual grounding approach

Using computer vision and deep learning (e.g., Convolutional Neural Networks) to automatically recognise unsafe behaviour from digital images can help managers identify and respond quickly to such actions and mitigate an adverse event. However, there has been a tendency for computer vision studies in construction to focus solely on detecting unsafe behaviour (i.e., object detection) or the regions of interest with pre-defined labels. Moreover, such approaches have been unable to consider rich semantic information among multiple unsafe actions in a digital image. The research we present in this paper uses a safety rule query to determine and locate several unsafe behaviours in a digital image by employing a visual grounding approach. Our approach consists of: (1) visual and text feature extraction, (2) recursive sub-query, and (3) generation of the bounding box. We validate our approach by conducting an experiment to demonstrate it is effectiveness. The results from an experimental study demonstrate an average precision, recall, and F1-score were 0.55, 0.85, and 0.65, respectively, suggesting our approach can accurately identify and locate different types of unsafe behaviours from digital images acquired from a construction site.     

Tangential Gap Flow (TGF) navigation: A new reactive obstacle avoidance approach for highly cluttered environments

This paper presents a novel reactive collision avoidance method for mobile robots moving in dense and cluttered environments. The proposed method, entitled Tangential Gap flow (TGF), simplifies the navigation problem using a divide and conquer strategy inspired by the well-known Nearness-Diagram Navigation (ND) techniques. At each control cycle, the TGF extracts free openings surrounding the robot and identifies the suitable heading which makes the best progress towards the goal. This heading is then adjusted to avoid the risk of collision with nearby obstacles based on two concepts namely, tangential and gap flow navigation. The tangential navigation steers the robot parallel to the boundary of the closest obstacle while still emphasizing the progress towards the goal. The gap flow navigation safely and smoothly drives the robot towards the free area in between obstacles that lead to the target. The resultant trajectory is faster, shorter and less-oscillatory when compared to the ND methods. Furthermore, identifying the avoidance maneuver is extended to consider all nearby obstacle points and generate an avoidance rule applicable for all obstacle configurations. Consequently, a smoother yet much more stable behavior is achieved. The stability of the motion controller, that guides the robot towards the desired goal, is proved in the Lyapunov sense. Experimental results including a performance evaluation in very dense and complex environments demonstrate the power of the proposed approach. Additionally, a discussion and comparison with existing Nearness-Diagram Navigation variants is presented.