論文

＜文献概要＞脳の計算理論の発展に伴い,それを精神疾患の病態理解に応用する試みへの期待が,近年,高まっている。特に,脳の予測符号化モデルに基づいた研究では,感覚情報の不確実性を推定する神経機能の変調と精神疾患との関連を示唆する概念的な仮説が提示されてきたが,実際の臨床で観察される行動レベルでの病像との間にはギャップが存在している。本論文では,神経ロボティクスの手法を用いることで,実験的にこの理論と臨床像との橋渡しを行った研究を紹介する。実験では,自閉スペクトラム症に特徴的な限定された反復的な行動様式や行為の停止といった多様な症状が,不確実性の過小評価と過大評価による異なるプロセスから生じ得ることが示され,精神症状のmultifinal性とequifinal性の一側面が捉えられた。また,病態理解にとどまらず,症状改善のための環境調整的な介入への理論的示唆といった,計算論的精神医学の臨床研究への貢献可能性が示された。

To work cooperatively with humans by using language, robots must not only acquire a mapping between language and their behavior but also autonomously utilize the mapping in appropriate contexts of interactive tasks online. To this end, we propose a novel learning method linking language to robot behavior by means of a recurrent neural network. In this method, the network learns from correct examples of the imposed task that are given not as explicitly separated sets of language and behavior but as sequential data constructed from the actual temporal flow of the task. By doing this, the internal dynamics of the network models both language behavior relationships and the temporal patterns of interaction. Here, "internal dynamics" refers to the time development of the system defined on the fixed-dimensional space of the internal states of the context layer. Thus, in the execution phase, by constantly representing where in the interaction context it is as its current state, the network autonomously switches between recognition and generation phases without any explicit signs and utilizes the acquired mapping in appropriate contexts. To evaluate our method, we conducted an experiment in which a robot generates appropriate behavior responding to a human's linguistic instruction. After learning, the network actually formed the attractor structure representing both language behavior relationships and the task's temporal pattern in its internal dynamics. In the dynamics, language behavior mapping was achieved by the branching structure. Repetition of human's instruction and robot's behavioral response was represented as the cyclic structure, and besides, waiting to a subsequent instruction was represented as the fixed-point attractor. Thanks to this structure, the robot was able to interact online with a human concerning the given task by autonomously switching phases.

Meanings of language expressions are constructed not only from words grounded in real-world matters, but also from words such as "not" that participate in the construction by working as logical operators. This study proposes a connectionist method for learning and internally representing functions that deal with both of these word groups, and grounding sentences constructed from them in corresponding behaviors just by experiencing raw sequential data of an imposed task. In the experiment, a robot implemented with a recurrent neural network is required to ground imperative positive and negative sentences given as a sequence of words in corresponding goal-oriented behavior. Analysis of the internal representations reveals that the network fulfilled the requirement by extracting XOR problems implicitly included in the target sequences and solving them by learning to represent the logical operations in its nonlinear dynamics in a self-organizing manner.

This study demonstrates that the prediction error minimization (PEM) mechanism can account for the emergence of reciprocal interaction between two cognitive agents. During interactive processes, alternation of forming and deforming interactions may be triggered by various internal and external causes. We focus in particular on external causes derived from a dynamic and uncertain environment. Two small humanoid robots controlled by an identical dynamic neural network model using the PEM mechanism were trained to achieve a set of coherent ball-playing interactions between them. The two robots predict each other in a top-down way while they try to minimize the prediction errors derived from the unstable ball dynamics or the external cause in a bottom-up way by using the PEM mechanism. The experimental results showed that switching among the set of trained interactive ball plays between the two robots appears spontaneously. The analysis clarified how each complementary behavior can be generated via mutual adaptation between the two robots by undertaking top-down and bottom-up interaction in each individual dynamic neural network model by using the PEM mechanism.

Traffic complexity is one of the factors affecting driver workload. In order to study the relationship between traffic complexity levels and workload, a designed experiment is required, especially to vary traffic flow parameters systematically in a simulated environment. This paper describes the experimental design of a simulator study for developing a computational model to estimate the behavior of driver workload based on traffic complexity. Driving simulators allow creating and testing different traffic scenarios and manipulating independent variables to improve the quality of data, as compared to real world experiments. Physiological responses such as heart rate, skin conductance, and pupil size have been found to be related to workload. By adapting a data-driven method, we integrated electrocardiography sensors, electro-dermal activity sensors, and eye-tracker to acquire driver physiological signals and gaze information. Preliminary results show a positive correlation between traffic complexity levels and corresponding physiological responses, performance, and subjective measures.

This paper investigates the essential difference between two types of behavior generation schemes, namely, sensory reflex behavior generation and intentional proactive behavior generation, by proposing a dynamic neural network model referred to as stochastic multiple-timescale recurrent neural network (S-MTRNN). The proposed model was employed in an experiment involving robots learning to cooperate with others under the condition of potential unpredictability of the others' behaviors. The results of the learning experiment showed that sensory reflex behavior was generated by a self-organizing probabilistic prediction mechanism when the initial sensitivity characteristics in the network dynamics were not utilized in the learning process. In contrast, proactive behavior with a deterministic prediction mechanism was developed when the initial sensitivity was utilized. It was further shown that in situations where unexpected behaviors of others were observed, the behavioral context was re-situated by adaptation of the internal neural dynamics by means of simple sensory reflexes in the former case. In the latter case, the behavioral context was re-situated by error regression of the internal neural activity rather than by sensory reflex. The role of the top-down and bottom-up interactions in dealing with unexpected situations is discussed.

In the present study, we demonstrate the learning and recognition capabilities of our recently proposed recurrent neural network (RNN) model called stochastic continuous-time RNN (S-CTRNN). S-CTRNN can learn to predict not only the mean but also the variance of the next state of the learning targets. The network parameters consisting of weights, biases, and initial states of context neurons are optimized through maximum likelihood estimation (MLE) using the gradient descent method. First, we clarify the essential difference between the learning capabilities of conventional CTRNN and S-CTRNN by analyzing the results of a numerical experiment in which multiple fluctuating temporal patterns were used as training data, where the variance of the Gaussian noise varied among the patterns. Furthermore, we also show that the trained S-CTRNN can recognize given fluctuating patterns by inferring the initial states that can reproduce the patterns through the same MLE scheme as that used for network training. © 2014 Springer International Publishing Switzerland.

This paper discusses a possible neurodynamic mechanism that enables self-organization of two basic behavioral modes, namely a 'proactive mode' and a 'reactive mode,' and of autonomous switching between these modes depending on the situation. In the proactive mode, actions are generated based on an internal prediction, whereas in the reactive mode actions are generated in response to sensory inputs in unpredictable situations. In order to investigate how these two behavioral modes can be self-organized and how autonomous switching between the two modes can be achieved, we conducted neurorobotics experiments by using our recently developed dynamic neural network model that has a capability to learn to predict time-varying variance of the observable variables. In a set of robot experiments under various conditions, the robot was required to imitate other's movements consisting of alternating predictable and unpredictable patterns. The experimental results showed that the robot controlled by the neural network model was able to proactively imitate predictable patterns and reactively follow unpredictable patterns by autonomously switching its behavioral modes. Our analysis revealed that variance prediction mechanism can lead to self-organization of these abilities with sufficient robustness and generalization capabilities.

We suggest that different behavior generation schemes, such as sensory reflex behavior and intentional proactive behavior, can be developed by a newly proposed dynamic neural network model, named stochastic multiple timescale recurrent neural network (S-MTRNN). The model learns to predict subsequent sensory inputs, generating both their means and their uncertainty levels in terms of variance (or inverse precision) by utilizing its multiple timescale property. This model was employed in robotics learning experiments in which one robot controlled by the S-MTRNN was required to interact with another robot under the condition of uncertainty about the other's behavior. The experimental results show that self-organized and sensory reflex behavior-based on probabilistic prediction-emerges when learning proceeds without a precise specification of initial conditions. In contrast, intentional proactive behavior with deterministic predictions emerges when precise initial conditions are available. The results also showed that, in situations where unanticipated behavior of the other robot was perceived, the behavioral context was revised adequately by adaptation of the internal neural dynamics to respond to sensory inputs during sensory reflex behavior generation. On the other hand, during intentional proactive behavior generation, an error regression scheme by which the internal neural activity was modified in the direction of minimizing prediction errors was needed for adequately revising the behavioral context. These results indicate that two different ways of treating uncertainty about perceptual events in learning, namely, probabilistic modeling and deterministic modeling, contribute to the development of different dynamic neuronal structures governing the two types of behavior generation schemes.

This study shows that a novel type of recurrent neural network model can learn to reproduce fluctuating training sequences by inferring their stochastic structures. The network learns to predict not only the mean of the next input state, but also its time-varying variance. The network is trained through maximum likelihood estimation by utilizing the gradient descent method, and the likelihood function is expressed as a function of both the predicted mean and variance. In a numerical experiment, in order to evaluate the performance of the model, we first tested its ability to reproduce fluctuating training sequences generated by a known dynamical system that were perturbed by Gaussian noise with state-dependent variance. Our analysis showed that the network can reproduce the sequences by predicting the variance correctly. Furthermore, the other experiment showed that a humanoid robot equipped with the network can learn to reproduce fluctuating tutoring sequences by inferring latent stochastic structures hidden in the sequences.

This study proposes a novel type of dynamic neural network model that can learn to extract stochastic or fluctuating structures hidden in time series data. The network learns to predict not only the mean of the next input state, but also its time-dependent variance. The training method is based on maximum likelihood estimation by using the gradient descent method and the likelihood function is expressed as a function of the estimated variance. Regarding the model evaluation, we present numerical experiments in which training data were generated in different ways utilizing Gaussian noise. Our analysis showed that the network can predict the time-dependent variance and the mean and it can also reproduce the target stochastic sequence data by utilizing the estimated variance. Furthermore, it was shown that a humanoid robot using the proposed network can learn to reproduce latent stochastic structures hidden in fluctuating tutoring trajectories. This learning scheme is essential for the acquisition of sensory-guided skilled behavior. © 2013 IEEE.

In this paper, we propose a method to dynamically modulate the input state of recurrent neural networks (RNNs) so as to realize flexible and robust robot behavior. We employ the so-called stochastic continuous-time RNN (S-CTRNN), which can learn to predict the mean and variance (or uncertainty) of subsequent sensorimotor information. Our proposed method uses this estimated uncertainty to determine a mixture ratio for combining actual and predicted sensory states of network input. The method is evaluated by conducting a robot learning experiment in which a robot is required to perform a sensory-dependent task and a sensory-independent task. The sensory-dependent task requires the robot to incorporate meaningful sensory information, and the sensory-independent task requires the robot to ignore irrelevant sensory information. Experimental results demonstrate that a robot controlled by our proposed method exhibits flexible and robust behavior, which results from dynamic modulation of the network input on the basis of the estimated uncertainty of actual sensory states.

Dynamic interactions with caregivers are essential for infants to develop cognitive abilities, including aspects of action, perception, and attention. We hypothesized that these abilities can be acquired through the predictive learning of sensory inputs including their uncertainty (inverse precision) in terms of variance. To examine our hypothesis from the perspective of cognitive developmental robotics, we conducted a neurorobotics experiment involving a ball-playing interaction task between a human experimenter representing a caregiver and a small humanoid robot representing an infant. The robot was equipped with a dynamic generative model called a stochastic continuous-time recurrent neural network (S-CTRNN). The S-CTRNN learned to generate predictions about both the visuo-proprioceptive states of the robot and the uncertainty of these states by minimizing a negative log-likelihood consisting of log-uncertainty and precision-weighted prediction error. The experimental results showed that predictive learning with uncertainty estimation enabled the robot to acquire infant-like cognitive abilities through dynamic interactions with the experimenter. We also discuss the effects of infant-directed modifications observed in caregiver-infant interactions on the development of these abilities.

An important characteristic of human language is compositionality. We can efficiently express a wide variety of real-world situations, events, and behaviors by compositionally constructing the meaning of a complex expression from a finite number of elements. Previous studies have analyzed how machine-learning models, particularly neural networks, can learn from experience to represent compositional relationships between language and robot actions with the aim of understanding the symbol grounding structure and achieving intelligent communicative agents. Such studies have mainly dealt with the words (nouns, adjectives, and verbs) that directly refer to real-world matters. In addition to these words, the current study deals with logic words, such as "not," "and," and "or" simultaneously. These words are not directly referring to the real world, but are logical operators that contribute to the construction of meaning in sentences. In human-robot communication, these words may be used often. The current study builds a recurrent neural network model with long short-term memory units and trains it to learn to translate sentences including logic words into robot actions. We investigate what kind of compositional representations, which mediate sentences and robot actions, emerge as the network's internal states via the learning process. Analysis after learning shows that referential words are merged with visual information and the robot's own current state, and the logical words are represented by the model in accordance with their functions as logical operators. Words such as "true," "false," and "not" work as non-linear transformations to encode orthogonal phrases into the same area in a memory cell state space. The word "and," which required a robot to lift up both its hands, worked as if it was a universal quantifier. The word "or," which required action generation that looked apparently random, was represented as an unstable space of the network's dynamical system.

In this paper, we propose a model that can explain the mechanism of self and non-self discrimination. Infants gradually develop their abilities for self-other cognition through interaction with the environment. Predictive learning has been widely used to explain the mechanism of infants' development. We hypothesized that infants' cognitive abilities are developed through predictive learning and the uncertainty estimation of their sensory-motor inputs. We chose a stochastic continuous time recurrent neural network, which is a dynamical neural network model, to predict uncertainties as variances. From the perspective of cognitive developmental robotics, a predictive learning experiment with a robot was performed. The results indicate that training made the robot predict the regions related to its body more easily. We confirmed that self and non-self cognitive abilities might be acquired through predictive learning with uncertainty estimation.

We propose a model for robots to use tools without predetermined parameters based on a human cognitive model. Almost all existing studies of robot using tool require predetermined motions and tool features, so the motion patterns are limited and the robots cannot use new tools. Other studies use a full search for new tools however, this entails an enormous number of calculations. We built a model for tool use based on the phenomenon of tool-body assimilation using the following approach: We used a humanoid robot model to generate random motion, based on human body babbling. These rich motion experiences were then used to train a recurrent neural network for modeling a body image. Tool features were self-organized in the parametric bias modulating the body image according to the used tool. Finally, we designed the neural network for the robot to generate motion only from the target image. © 2014 Springer International Publishing Switzerland.

This article presents a new data-driven framework for analyzing periodic physical human-robot interaction (pHRI) in latent state space. The model representing pHRI is critical for elaborating human understanding and/or robot control during pHRI. Recent advancements in deep learning technology would allow us to train such a model on a dataset collected from the actual pHRI. Our framework is based on a variational recurrent neural network (VRNN), which can process time-series data generated by a pHRI. This study modifies VRNN to explicitly integrate the latent dynamics from robot to human and to distinguish it from a human state estimate module. Furthermore, to analyze periodic motions, such as walking, we integrate VRNN with a new recurrent network based on reservoir computing (RC), which has random and fixed connections between numerous neurons. By boosting RC into a complex domain, periodic behavior can be represented as phase rotation in the complex domain without decaying the amplitude. A rope rotation/swinging experiment was used to validate the proposed framework. The proposed framework, trained on the collected experiment dataset, achieved the latent state space in which variation in periodic motions can be distinguished. The best prediction accuracy of the human observations and robot actions was obtained in such a well-distinguished space.

In various domains, including production and daily life environments, there is growing demand for robots capable of manipulating flexible objects such as cables and fabrics. Reinforcement learning and imitation learning are two of the most promising approaches for realizing such robots; however, they face issues related to the high cost of data collection and labeling, respectively. To address these issues, we have shifted our focus to 'learning from play,' where play data refers to data collected by asking human operators to teleoperate a robot while maintaining a sense of curiosity. Specifically, we employ a framework called play-supervised latent motor plans (Play-LMP), which was originally evaluated for rigid object manipulation in a simulation environment. We extend this Play-LMP framework to the domain of flexible object manipulation in a real-world environment and evaluate its potential. Our experimental results on fixed-end rope manipulation demonstrate both the capabilities and limitations of the framework. In addition, we show that the latent plan space of the framework becomes structured according to specific tasks, even without explicit labeling of the data. These results suggest the effectiveness of our approach in flexible object manipulation because it achieves a reduction in data collection costs and enables task-unrestricted manipulations.

Long-horizon tasks require intelligent agents, such as robots, to handle both temporal uncertainty and temporal dependency. Although world models have shown promise for solving tasks across many domains, they often struggle with managing long context in tasks due to the limited representation ability of their latent dynamics models. To overcome this issue, we propose a novel hierarchical latent dynamics model that takes into account multiple-timescale dynamics. Specifically, our proposed model, called the 'multiple timescale recurrent state-space model' (MTRSSM), comprises a higher level with slow dynamics and a lower level with fast dynamics, each incorporating both deterministic and stochastic latent states. We demonstrate, both quantitatively and qualitatively, that a world model with our proposed MTRSSM can generate superior video predictions for long-horizon robotic object-manipulation tasks through latent imagination compared with other baselines. Importantly, we emphasize the critical role of the higher level in effectively handling temporal uncertainty and temporal dependency in long-horizon tasks. These findings indicate that the proposed MTRSSM enables intelligent agents to acquire a better understanding of the environment and generate more accurate predictions, thereby facilitating their learning and planning of long-horizon tasks.

Although psychiatric disorders have conventionally been treated categorically, recent research indicates a continuous and hierarchical structure among psychiatric symptoms, with a general psychopathology factor (p-factor) at the top and several specific factors at lower levels. However, understanding of the psychometric properties of these factors in the general population remains limited. In this study, analyzing responses from approximately 1000 individuals across three datasets to various diagnostic questionnaires, we extracted general and specific factors using diverse hierarchical structures and factor numbers. Comparative analysis demonstrated robust properties of the general factor across different models, with higher explained variance and internal consistency compared with specific factors. Consistent with the p-factor concept, the general factor was correlated with diverse diagnostic histories. Conversely, the content of specific factors varied with factor numbers, and some corresponded to traditional diagnostic categories. Our findings, we believe, provide a scaffold for disentangling the complex structure among psychiatric symptoms.