Measuring generalization of visuomotor perturbations in wrist movements using mobile phones

Recent studies in motor control have shown that visuomotor rotations for reaching have narrow generalization functions: what we learn during movements in one direction only affects subsequent movements into close directions. Here we wanted to measure the generalization functions for wrist movement. To do so we had 7 subjects performing an experiment holding a mobile phone in their dominant hand. The mobile phone's built in acceleration sensor provided a convenient way to measure wrist movements and to run the behavioral protocol. Subjects moved a cursor on the screen by tilting the phone. Movements on the screen toward the training target were rotated and we then measured how learning of the rotation in the training direction affected subsequent movements in other directions. We find that generalization is local and similar to generalization patterns of visuomotor rotation for reaching.     

Role of uncertainty in sensorimotor control

Neural signals are corrupted by noise and this places limits on information processing. We review the processes involved in goal-directed movements and how neural noise and uncertainty determine aspects of our behaviour. First, noise in sensory signals limits perception. We show that, when localizing our hand, the central nervous system (CNS) integrates visual and proprioceptive information, each with different noise properties, in a way that minimizes the uncertainty in the overall estimate. Second, noise in motor commands leads to inaccurate movements. We review an optimal-control framework, known as 'task optimization in the presence of signal-dependent noise', which assumes that movements are planned so as to minimize the deleterious consequences of noise and thereby minimize inaccuracy. Third, during movement, sensory and motor signals have to be integrated to allow estimation of the body's state. Models are presented that show how these signals are optimally combined. Finally, we review how the CNS deals with noise at the neural and network levels. In all of these processes, the CNS carries out the tasks in such a way that the detrimental effects of noise are minimized. This shows that it is important to consider effects at the neural level in order to understand performance at the behavioural level.     

Neural-space generalization of a topological transformation

An investigation is performed to assess the generalization capability found in neural network paradigms to approximate a 2-dimensional coordinate (topological) transformation. A developed strategy uses the example to give a physical meaning to what is meant by generalization. The example shows how to use a neural network paradigm to generalize a two-degree of freedom topological transformation from cartesian end-point coordinates to corresponding joint angle coordinates based only on examples of the mapping. The importance of this example is that it provides a clear understanding of how and what a neural network is actually communications and brings a theoretical idea to a useful understanding. When examples characterize the topology, a collective generalization property begins to emerge and the network learns the topology. If the network is presented with additional examples of the transformation, the network can generate the corresponding joint angles to any unseen position, that is, by generalization. It is also significant that the network's generalization property emerges from the network based on very few training examples! Further, the networks power exists with very few neurons. Results suggest the use of the paradigm's generalization capability to provide solutions to unknown or intractable algorithms for applications.     

Tuning diversity in bagged ensembles

In this paper, we investigate how the level of diversity amongst individual neural networks in a bagged ensemble can significantly influence overall ensemble generalization performance. We propose a new technique that tunes this diversity so that ensemble generalization performance is optimized and evaluate its performance on benchmark regression data-sets.     

Cortical mechanisms of sensory learning and object recognition

Learning about the world through our senses constrains our ability to recognise our surroundings. Experience shapes perception. What is the neural basis for object recognition and how are learning-induced changes in recognition manifested in neural populations? We consider first the location of neurons that appear to be critical for object recognition, before describing what is known about their function. Two complementary processes of object recognition are considered: discrimination among diagnostic object features and generalization across non-diagnostic features. Neural plasticity appears to underlie the development of discrimination and generalization for a given set of features, though tracking these changes directly over the course of learning has remained an elusive task.     

Generalization of motor learning depends on the history of prior action

Generalization of motor learning refers to our ability to apply what has been learned in one context to other contexts. When generalization is beneficial, it is termed transfer, and when it is detrimental, it is termed interference. Insight into the mechanism of generalization may be acquired from understanding why training transfers in some contexts but not others. However, identifying relevant contextual cues has proven surprisingly difficult, perhaps because the search has mainly been for cues that are explicit. We hypothesized instead that a relevant contextual cue is an implicit memory of action with a particular body part. To test this hypothesis we considered a task in which participants learned to control motion of a cursor under visuomotor rotation in two contexts: by moving their hand through motion of their shoulder and elbow, or through motion of their wrist. Use of these contextual cues led to three observations: First, in naive participants, learning in the wrist context was much faster than in the arm context. Second, generalization was asymmetric so that arm training benefited subsequent wrist training, but not vice versa. Third, in people who had prior wrist training, generalization from the arm to the wrist was blocked. That is, prior wrist training appeared to prevent both the interference and transfer that subsequent arm training should have caused. To explain the data, we posited that the learner collected statistics of contextual history: all upper arm movements also move the hand, but occasionally we move our hands without moving the upper arm. In a Bayesian framework, history of limb segment use strongly affects parameter uncertainty, which is a measure of the covariance of the contextual cues. This simple Bayesian prior dictated a generalization pattern that largely reproduced all three findings. For motor learning, generalization depends on context, which is determined by the statistics of how we have previously used the various parts of our limbs.     

Encoding time in neural dynamic regimes with distinct computational tradeoffs

Converging evidence suggests the brain encodes time in dynamic patterns of neural activity, including neural sequences, ramping activity, and complex dynamics. Most temporal tasks, however, require more than just encoding time, and can have distinct computational requirements including the need to exhibit temporal scaling, generalize to novel contexts, or robustness to noise. It is not known how neural circuits can encode time and satisfy distinct computational requirements, nor is it known whether similar patterns of neural activity at the population level can exhibit dramatically different computational or generalization properties. To begin to answer these questions, we trained RNNs on two timing tasks based on behavioral studies. The tasks had different input structures but required producing identically timed output patterns. Using a novel framework we quantified whether RNNs encoded two intervals using either of three different timing strategies: scaling, absolute, or stimulus-specific dynamics. We found that similar neural dynamic patterns at the level of single intervals, could exhibit fundamentally different properties, including, generalization, the connectivity structure of the trained networks, and the contribution of excitatory and inhibitory neurons. Critically, depending on the task structure RNNs were better suited for generalization or robustness to noise. Further analysis revealed different connection patterns underlying the different regimes. Our results predict that apparently similar neural dynamic patterns at the population level (e.g., neural sequences) can exhibit fundamentally different computational properties in regards to their ability to generalize to novel stimuli and their robustness to noise—and that these differences are associated with differences in network connectivity and distinct contributions of excitatory and inhibitory neurons. We also predict that the task structure used in different experimental studies accounts for some of the experimentally observed variability in how networks encode time.     

Tradeoff between stability and maneuverability during whole-body movements

Background

Understanding how stability and/or maneuverability affects motor control strategies can provide insight on moving about safely in an unpredictable world. Stability in human movement has been well-studied while maneuverability has not. Further, a tradeoff between stability and maneuverability during movement seems apparent, yet has not been quantified. We proposed that greater maneuverability, the ability to rapidly and purposefully change movement direction and speed, is beneficial in uncertain environments. We also hypothesized that gaining maneuverability comes at the expense of stability and perhaps also corresponds with decreased muscle coactivation.

Materials and Methods

We used a goal-directed forward lean movement task that integrated both stability and maneuverability. Subjects (n = 11) used their center of pressure to control a cursor on a computer monitor to reach a target. We added task uncertainty by shifting the target anterior-posterior position mid-movement. We used a balance board with a narrow beam that reduced the base of support in the medio-lateral direction and defined stability as the probability that subjects could keep the balance board level during the task.

Results

During the uncertainty condition, subjects were able to change direction of their anterior-posterior center of pressure more rapidly, indicating that subjects were more maneuverable. Furthermore, medio-lateral center of pressure excursions also approached the edges of the beam and reduced stability margins, implying that subjects were less stable (i.e. less able to keep the board level). On the narrow beam board, subjects increased muscle coactivation of lateral muscle pairs and had greater muscle activity in the left leg. However, there were no statistically significant differences in muscle activity amplitudes or coactivation with uncertainty.

Conclusions/Significance

These results demonstrate that there is a tradeoff between stability and maneuverability during a goal-directed whole-body movement. Tasks with added uncertainty could help individuals learn to be more maneuverable yet sufficiently stable.     

A review of the generalization of auditory learning

The ability to detect and discriminate attributes of sounds improves with practice. Determining how such auditory learning generalizes to stimuli and tasks that are not encountered during training can guide the development of training regimens used to improve hearing abilities in particular populations as well as provide insight into the neural mechanisms mediating auditory performance. Here we review the newly emerging literature on the generalization of auditory learning, focusing on behavioural investigations of generalization on basic auditory tasks in human listeners. The review reveals a variety of generalization patterns across different trained tasks that can not be summarized with a simple rule, and a diversity of views about the definition, evaluation and interpretation of generalization.     

Generalization and the evolution of symmetry preferences

Biological displays are often symmetrical, and there is growing evidence that receivers are sensitive to these symmetries. One explanation for the evolution of such sensitivity is that symmetry reflects the quality of the signaller. An alternative is that the sensitivity may arise as a by-product of general properties of biological recognition systems. In line with the latter idea, simulations of the recognition process based on simple, artificial neural networks have suggested that generalization can give rise to preferences for particular symmetrical stimuli. However, it is not clear from these studies exactly how the preferences emerge, and to what extent the results are relevant to biological recognition systems. Here, we employ a different class of recognition models (gradient interaction models) to demonstrate more clearly how generalization can generate a preference for symmetrical variants of a display. We also point out that the predictions of the gradient interaction and network-based models regarding the effects of generalization closely match the results from empirical studies of stimulus control. Our analysis demonstrates that the effects of generalization cannot be ignored when studying the evolution of symmetry preferences and symmetric signals.     

运动对时距知觉的影响及其神经机制

在运动过程中,时距知觉的能力非常重要,能帮助个体对时长进行判断及对事件的发生做出预测和准备.近年来,越来越多的研究发现运动本身会直接影响个体的时距知觉.本文分别从运动参数、运动阶段、视觉运动刺激和运动有关的个体因素四个方面梳理了运动对时距知觉产生影响的行为学证据.目前已经有大量研究从不同角度证明,大脑运动系统组成了支持...     

Moving, sensing and learning with cerebellar damage

The cerebellum is a subcortical brain structure that is essential for learning and controlling movement. Recent work shows that the cerebellum also plays a role in certain perceptual abilities, beyond what would be expected secondary to poor movement control. This review covers these and other recent advances, focusing on how cerebellar damage affects human abilities ranging from sensory perception to movement control and motor learning.     

Form and function in systems neuroscience

'Form follows function' is an architectural philosophy attributed to the great American architect Louis Sullivan, and later taken up by the Bauhaus movement. It stresses that the form of a building should reflect its function. Neuroscientists have used the converse of this dictum to learn the functions of neural circuits, believing that if we study neural architecture, it will lead us to an understanding of how neural systems function. New tools for studying the structure of neural circuits are being developed, so it is important to discuss what the old techniques have taught us about how to derive function from the form of a neural circuit.     

Neural Tuning Functions Underlie Both Generalization and Interference

In sports, the role of backswing is considered critical for generating a good shot, even though it plays no direct role in hitting the ball. We recently demonstrated the scientific basis of this phenomenon by showing that immediate past movement affects the learning and recall of motor memories. This effect occurred regardless of whether the past contextual movement was performed actively, passively, or shown visually. In force field studies, it has been shown that motor memories generalize locally and that the level of compensation decays as a function of movement angle away from the trained movement. Here we examine if the contextual effect of past movement exhibits similar patterns of generalization and whether it can explain behavior seen in interference studies. Using a single force-field learning task, the directional tuning curves of both the prior contextual movement and the subsequent force field adaptive movements were measured. The adaptation movement direction showed strong directional tuning, decaying to zero by 90° relative to the training direction. The contextual movement direction exhibited a similar directional tuning, although the effect was always above 60%. We then investigated the directional tuning of the passive contextual movement using interference tasks, where the contextual movements that uniquely specified the force field direction were separated by ±15° or ±45°. Both groups showed a pronounced tuning effect, which could be well explained by the directional tuning functions for single force fields. Our results show that contextual effect of past movement influences predictive force compensation, even when adaptation does not require contextual information. However, when such past movement contextual information is crucial to the task, such as in an interference study, it plays a strong role in motor memory learning and recall. This work demonstrates that similar tuning responses underlie both generalization of movement direction during dynamic learning and contextual movements in both single force field and interference tasks.     

The neural circuits and synaptic mechanisms underlying motor initiation in C. elegans

C. elegans is widely used to dissect how neural circuits and genes generate behavior. During locomotion, worms initiate backward movement to change locomotion direction spontaneously or in response to sensory cues; however, the underlying neural circuits are not well defined. We applied a multidisciplinary approach to map neural circuits in freely behaving worms by integrating functional imaging, optogenetic interrogation, genetic manipulation, laser ablation, and electrophysiology. We found that a disinhibitory circuit and a stimulatory circuit together promote initiation of backward movement and that circuitry dynamics is differentially regulated by sensory cues. Both circuits require glutamatergic transmission but depend on distinct glutamate receptors. This dual mode of motor initiation control is found in mammals, suggesting that distantly related organisms with anatomically distinct nervous systems may adopt similar strategies for motor control. Additionally, our studies illustrate how a multidisciplinary approach facilitates dissection of circuit and synaptic mechanisms underlying behavior in a genetic model organism.     

In search of certainty in revolutionary India

In this article, I focus on the dilemmas of a friend in Jharkhand who might have joined the revolutionary armed squads of the Maoist insurgency in India. I show how the question of why one supports a revolutionary movement, as well as the nature of that support, can change over time. In particular, I stress the importance of the dialectics between epistemological and ontological uncertainty and certainty, which may be central to the making of a revolutionary in a particular phase of a revolutionary movement. These dialectics are not just the result of an ontological uncertainty of ideological commitment to the movement, but are crucially also about the search for epistemological clarity in social relations imagined to be less opaque and hence more trustworthy. Doubt, an uncertainty about what one knows about one’s social relationships is characteristic of the epistemic murk that accompanies the breakdown of the normative order in the revolutionary situation analysed here. In this context, Maoist terror arises from the creation of epistemic clarity—the possibility that on the other side norms and relationships will be more certain. This is a certainty that is carved out of uncertainty and ambivalence, a certainty that denies or projects away uncertainty. Its weapon is paranoia, an ability to make enemies where there would be doubt, betrayal where there would be benefit of the doubt. The potential revolutionary is therefore not only unsure about his/her ideological commitments, but moreover, a crucial component of their predicament might be an uncertainty about the social relations in which they find themselves and the hope that revolutionary engagement might come with more guarantees. Becoming a revolutionary is also about being in search of certainty.     

Substantial Generalization of Sensorimotor Learning from Bilateral to Unilateral Movement Conditions

Controversy exists regarding whether bimanual skill learning can generalize to unimanual performance. For example, some investigators showed that dynamic adaptation could only partially generalize between bilateral and unilateral movement conditions, while others demonstrated complete generalization of visuomotor adaptation. Here, we identified three potential factors that might have contributed to the discrepancy between the two sets of findings. In our first experiment, subjects performed reaching movements toward eight targets bilaterally with a novel force field applied to both arms, then unilaterally with the force field applied to one arm. Results showed that the dynamic adaptation generalized completely from bilateral to unilateral movements. In our second experiment, the same force field was only applied to one arm during both bilateral and unilateral movements. Results indicated complete transfer again. Finally, our subjects performed reaching movements toward a single target with the force field or a novel visuomotor rotation applied only to one arm during both bilateral and unilateral movements. The reduced breadth of experience obtained during bilateral movements resulted in incomplete transfer, which explains previous findings of limited generalization. These findings collectively suggest a substantial overlap between the neural processes underlying bilateral and unilateral movements, supporting the idea that bilateral training, often employed in stroke rehabilitation, is a valid method for improving unilateral performance. However, our findings also suggest that while the neural representations developed during bilateral training can generalize to facilitate unilateral performance, the extent of generalization may depend on the breadth of experience obtained during bilateral training.     

A benchmark for how well neural nets generalize

This paper discusses the problem of extending the domain of learning sets and introduces HERBIE, a program which achieves this through graphical procedures rather than via neural networks. It is argued that for theoretical reasons HERBIE is well-suited to serving as a benchmark for measuring generalization efficacy, and therefore to serving as a means of testing claims of emergent distributed intelligence in neural nets. The successful results of tests of HERBIE as a pattern recognizer are presented, and HERBIE's behavior is favorably compared to neural nets for several real generalization problems. Finally, applications of HERBIE independent of its serving as a generalization benchmark, particularly in the area of cognitive science, are discussed.     

How robust are neural network models of stimulus generalization?

Artificial feed-forward neural networks are commonly used as a tool for modelling stimulus selection and animal signalling. A key finding of stimulus selection research has been generalization: if a given behaviour has been established to one stimulus, perceptually similar novel stimuli are likely to induce a similar response. Stimulus generalization, in feed-forward neural networks, automatically arises as a property of the network. This network property raises understandable concern regarding the sensitivity of the network to variation in its internal parameter values used in relation to its structure and to its training process. Researchers must have confidence that the predictions of their model follow from the underlying biology that they deliberately incorporated in the model, and not from often arbitrary choices about model implementation. We study how network training and parameter perturbations influence the qualitative and quantitative behaviour of a simple but general network. Specifically, for models of stimulus control we study the effect that parameter variation has on the shape of the generalization curves produced by the network. We show that certain network and training conditions produce undesirable artifacts that need to be avoided (or at least understood) when modelling stimulus selection.     

Spatial Generalization in Operant Learning: Lessons from Professional Basketball

In operant learning, behaviors are reinforced or inhibited in response to the consequences of similar actions taken in the past. However, because in natural environments the “same” situation never recurs, it is essential for the learner to decide what “similar” is so that he can generalize from experience in one state of the world to future actions in different states of the world. The computational principles underlying this generalization are poorly understood, in particular because natural environments are typically too complex to study quantitatively. In this paper we study the principles underlying generalization in operant learning of professional basketball players. In particular, we utilize detailed information about the spatial organization of shot locations to study how players adapt their attacking strategy in real time according to recent events in the game. To quantify this learning, we study how a make \ miss from one location in the court affects the probabilities of shooting from different locations. We show that generalization is not a spatially-local process, nor is governed by the difficulty of the shot. Rather, to a first approximation, players use a simplified binary representation of the court into 2 pt and 3 pt zones. This result indicates that rather than using low-level features, generalization is determined by high-level cognitive processes that incorporate the abstract rules of the game.