Search

Central pattern generators (CPGs) form the neural basis of the observed rhythmic behaviors for locomotion in legged animals. The CPG dynamics organized into networks allow the emergence of complex locomotor behaviors. In this work, we take this inspiration for developing walking behaviors in multilegged robots. We present novel DeepCPG policies that embed CPGs as a layer in a larger neural network and facilitate end-to-end learning of locomotion behaviors in deep reinforcement learning (DRL) setup. We demonstrate the effectiveness of this approach on physics engine-based insectoid robots. We show that, compared to traditional approaches, DeepCPG policies allow sample-efficient end-to-end learning of effective locomotion strategies even in the case of high-dimensional sensor spaces (vision). We scale the DeepCPG policies using a modular robot configuration and multiagent DRL. Our results suggest that gradual complexification with embedded priors of these policies in a modular fashion could achieve nontrivial sensor and motor integration on a robot platform. These results also indicate the efficacy of bootstrapping more complex intelligent systems from simpler ones based on biological principles. Finally, we present the experimental results for a proof-of-concept insectoid robot system for which DeepCPG learned policies initially using the simulation engine and these were afterward transferred to real-world robots without any additional fine-tuning.

Deep reinforcement learning (RL) has made it possible to solve complex robotics problems using neural networks as function approximators. However, the policies trained on stationary environments suffer in terms of generalization when transferred from one environment to another. In this work, we use Robust Markov Decision Processes (RMDP) to train the drone control policy, which combines ideas from Robust Control and RL. It opts for pessimistic optimization to handle potential gaps between policy transfer from one environment to another. The trained control policy is tested on the task of quadcopter positional control. RL agents were trained in a MuJoCo simulator. During testing, different environment parameters (unseen during the training) were used to validate the robustness of the trained policy for transfer from one environment to another. The robust policy outperformed the standard agents in these environments, suggesting that the added robustness increases generality and can adapt to nonstationary environments.

“Happy New Year!” At the beginning of 2016, I would like to take this opportunity to wish everyone a very happy, healthy, and prosperous new year! It is my great honor and privilege to serve as the Editor-in-Chief (EiC) of the IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (TNNLS), and I am excited to write this Editorial to start a new journey with you all.

Attractor networks have been one of the most successful paradigms in neural computation, and have been used as models of computation in the nervous system. Recently, we proposed a paradigm called ‘‘latent attractors’’ where attractors embedded in a recurrent network via Hebbian learning are used to channelnetwork response to external input rather than becoming manifest themselves. This allows the network to generate context-sensitive internal codes in complex situations. Latent attractors are particularly helpful in explaining computations within the hippocampus——a brain region of fundamental significance for memory and spatial learning.Latent attractor networks are a special case of associative memory networks. The model studied here consists of a two-layer recurrent network with attractors stored in the recurrent connections using a clipped Hebbian learning rule. The firing in both layers is competitive——Kwinners take all firing. The number of neurons allowed to fire, K, is smaller than the size of the active set of the stored attractors. The performance of latent attractor networks depends on the number of such attractors that a network can sustain.In this paper, we use signal-to-noise methods developed for standard associative memory networks to do a theoretical and computational analysis of the capacity and dynamics of latent attractor networks. This is an important first step in making latent attractors a viable tool in the repertoire of neural computation. The method developed here leads to numerical estimates of capacity limits and dynamics of latent attractor networks. The technique represents a general approach to analyse standard associative memory networks with competitive firing.The theoretical analysis is based on estimates of the dendritic sum distributions using Gaussian approximation. Because of the competitive firing property, the capacity results are estimated only numerically by iteratively computing the probability of erroneous firings. The analysis contains two cases: the simple case analysis which accounts for the correlations between weights due to shared patterns and the detailed case analysis which includes also the temporal correlations between the network's present and previous state. The latter case predicts better the dynamics of the network state for non-zero initial spurious firing. The theoretical analysis also shows the influence of the main parameters of the model on the storage capacity.

Recently, it has been suggested that attractor networks may provide a mechanism for context-dependence in hippocampal place codes. We have proposed that context may be coded by “latent attractors” — mutually competitive and internally cooperative cell groups which channel the system's response to afferent stimuli. We have also argued that it is the disynaptically recurrent dentate gyrus–hilus (DGH) system which embodies these latent attractors in the hippocampus. Others have suggested the CA3 network — with its monosynaptic recurrence — is the site. While latent attractors (and, thus, context-dependent coding) can be implemented by a 1-layer or 2-layer recurrent network, we show that a 2-layer recurrent network can implement a sparse context-dependent place-code using latent attractors much more flexibly than a 1-layer network.

Recurrent neural networks with full symmetric connectivity have been extensively studied as associative memories and pattern recognition devices. However, there is considerable evidence that sparse, asymmetrically connected, mainly excitatory networks with broadly directed inhibition are more consistent with biological reality. In this paper, we use the technique of return maps to study the dynamics of random networks with sparse, asymmetric connectivity and nonspecific inhibition. These networks show three qualitatively different kinds of behavior: fixed points, cycles of low period, and extremely long cycles verging on aperiodicity. Using statistical arguments, we relate these behaviors to network parameters and present empirical evidence for the accuracy of this statistical model. The model, in turn, leads to methods for controlling the level of activity in networks. Studying random, untrained networks provides an understanding of the intrinsic dynamics of these systems. Such dynamics could provide a substrate for the much more complex behavior shown when synaptic modification is allowed.

We investigate the dynamics of a class of recurrent random networks with sparse, asymmetric excitatory connectivity and global shunting inhibition mediated by a single interneuron. Using probabilistic arguments and a hyperbolic tangent approximation to the gaussian, we develop a simple method for setting the average level of firing activity in these networks. We demonstrate through simulations that our technique works well and extends to networks with more complicated inhibitory schemes. We are interested primarily in the CA3 region of the mammalian hippocampus, and the random networks investigated here are seen as modeling the a priori dynamics of activity in this region. In the presence of external stimuli, a suitable synaptic modification rule could shape this dynamics to perform temporal information processing tasks such as sequence completion and prediction.