V Srinivasa Chakravarthy

A computational model of the Basal Ganglia and the Hippocampus as key players in solving a navigation task is presented. The roles played by the above-mentioned neural substrates in navigation are demonstrated by an exploration task performed by a model rat in a simulated Morris Water Maze. To highlight the role of hippocampus in navigation, the agent is made to adopt a context-based navigation strategy. To demonstrate the role of BG in navigation, the agent is made to adopt a visual cue-based navigation strategy. The models are developed based on "actor-critic" architecture and trained using reinforcement learning. The above two models are integrated into a complete model which incorporates the above two forms of navigation. © 2010 Springer-Verlag Berlin Heidelberg.

We develop a sequential Q-learning model using a recurrent neural network to count objects in images using attentional search. The proposed model, which is based on visual attention, scans images by making a sequence of attentional jumps or saccades. By integrating the information gathered by the sequence of saccades, the model counts the number of targets in the image. The model consists primarily of two modules: the Classification Network and the Saccade Network. Whereas the Classification network predicts the number of target objects in the image, the Saccade network predicts the next saccadic jump. When the probability of the best predicted class crosses a threshold, the model halts making saccades and outputs its class prediction. Correct prediction results in positive reward, which is used to train the model by Q-learning. We achieve an accuracy of 92.1% in object counting. Simulations show that there is a direct relation between the number of glimpses required and the number of objects present to achieve a high accuracy in object counting.

A large-scale model of brain dynamics, as it is manifested in functional neuroimaging data, is presented in this study. The model is built around a general trainable network of Hopf oscillators, the dynamics of which are described in the complex domain. It was shown earlier that when a pair of Hopf oscillators are coupled by power coupling with a complex coupling strength, it is possible to stabilize the normal phase difference at a value related to the angle of the complex coupling strength. In the present model, the magnitudes of the complex coupling weights are set using the Structural Connectivity information obtained from Diffusion Tensor Imaging (DTI). The complex-valued outputs of the oscillator network are transformed by a complex-valued feedforward network with a single hidden layer. The entire model is trained in 2 stages: in the 1 st stage, the intrinsic frequencies of the oscillators in the oscillator network are trained, whereas in the 2 nd stage, the weights of the feedforward network are trained using the complex backpropagation algorithm. The Functional Connectivity Matrix (FCM) obtained from the network’s output is compared with empirical Functional Connectivity Matrix, a comparison that resulted in a correlation of 0.99 averaged over 5 subjects.

Sign Language (SL) is a potential tool for communication in the hearing and speech-impaired community. As individual words cannot be communicated accurately using the SL gestures, fingerspelling is adopted to spell out names of people and places. Due to rich vocabulary and diversity in Indic scripts, and the abugida nature of Indic scripts that distinguish them from a prominent world script like the Roman script, it is cumbersome to use American Sign Language (ASL) convention for fingerspelling in Indian languages. Moreover, due to the existence of 10 major scripts in India, it is a futile task to develop a separate fingerspelling convention for each individual Indic script based on the geometry of the characters. In this paper, we propose a novel and unified fingerspelling system known as Mudrabharati for Indic scripts. The gestures of Mudrabharati are constructed based on the phonetics of Indian scripts and not the geometry of the glyphs that compose the individual characters. Unlike ASL that utilizes just one hand, Mudrabharati uses both the hands - one for consonants and the other for vowels; swarayukta aksharas (Consonant-Vowel combinations) are gestured by using both the hands. An Artificial Intelligence (AI) based recognition system for Mudrabharati that returns the character in Devanagari and Tamil scripts is developed.

We present a model of basal ganglia as a key player in exploratory behavior. The model describes exploration of a virtual rat in a simulated "water pool" experiment. The virtual rat is trained using a reward-based or reinforcement learning paradigm which requires units with stochastic behavior for exploration of the system's state space. We model the STN-GPe system as a pair of neuronal layers with oscillatory dynamics, exhibiting a variety of dynamic regimes like chaos, traveling waves and clustering. Invoking the property of chaotic systems to explore a state space, we suggest that the complex "exploratory" dynamics of STN-GPe system in conjunction with dopamine-based reward signaling present the two key ingredients of a reinforcement learning system. © Springer-Verlag Berlin Heidelberg 2004.

We present a technique of using Deep Neural Networks (DNNs) for offline character recognition of Telugu characters. We construct DNNs by stacking Auto-encoders that are trained in a greedy layer-wise fashion in an unsupervised manner. We then perform supervised fine-tuning to train the entire network. We provide results on Consonant and Vowel Modifier Datasets using two and three hidden layer DNNs. We also construct an ensemble classifier to increase the classification performance further. We observe 94.25% accuracy for the two hidden layer network on Consonant data and 94.1% on Vowel Modifier Dataset which increases to 95.4% for Consonant and 94.8% for Vowel Modifier Dataset after combining classifiers to form an ensemble classifier of 4 different two hidden layer networks. © 2012 Springer-Verlag.

We present a model of bird song production in which the motor control pathway is modeled by a trainable network of oscillators and the Anterior Forebrain Pathway (AFP) is modeled as a stochastic system. We hypothesize 1) that the songbird learns only evaluations of songs during the sensory phase; 2) that the AFP plays a role analogous to the Explorer, a key component in Reinforcement Learning (RL); 3) the motor pathway learns the song by combining the evaluations (Value information) stored from the sensory phase, and the exploratory inputs from the AFP in a temporal stage-wise manner. Model performance from real birdsong samples is presented. © 2010 Springer-Verlag Berlin Heidelberg.

We present a model of basal ganglia (BG) that departs from the classical Go/NoGo picture of the function of its key pathways-the Direct and Indirect Pathways (DP and IP). Between the Go and NoGo regimes, we posit a third Explore regime, which denotes random exploration of action alternatives. Striatal dopamine (DA) is assumed to switch between DP and IP activation. The IP is modeled as a loop of the subthalamic nucleus (STN) and the Globus Pallidus externa (GPe). Simulations reveal that while the model displays Go and NoGo regimes for extreme values of DA, at intermediate values of DA, it exhibits exploratory behavior, which originates from the chaotic activity of the STN-GPe loop. We describe a series of BG models based on Go/Explore/NoGo approach, to explain the role of BG in three cases: (1) a simple action selection task, (2) reaching, and (3) willed action. © 2013 Elsevier B.V.

Making an optimal decision could be to either ‘Explore’ or ‘exploit’ or ‘not to take any action,’ and basal ganglia (BG) are considered to be a key neural substrate in decision making. In earlier chapters, we had hypothesized earlier that the indirect pathway (IP) of the BG could be the subcortical substrate for exploration. Here, we build a spiking network model to relate exploration to synchrony levels in the BG (which are a neural marker for tremor in Parkinson’s disease). Key BG nuclei such as the subthalamic nucleus (STN), Globus Pallidus externus (GPe), and Globus Pallidus internus (GPi) were modeled as Izhikevich spiking neurons, whereas the striatal output was modeled as Poisson spikes. We have applied reinforcement learning framework with the dopamine signal representing the reward prediction error used for cortico-striatal weight update. We apply the model to two decision-making tasks: a binary action selection task and an n-armed bandit task. The model shows that exploration levels could be controlled by STN’s lateral connection strength which also influenced the synchrony levels in the STN–GPe circuit. An increase in STN’s lateral strength led to a decrease in exploration which can be thought as the possible explanation for reduced exploratory levels in Parkinson’s patients.

The traditional approach to characterization of orientation maps as they were expounded by Hubel and Wiesel treats them as static representations. Only the magnitude of a neuron’s firing response to orientation is considered and the neuron with the highest response is said to be “tuned” to that response. But the neuronal response to orientation is a time-varying spike train and, if the response of an entire cortical area that potentially responds to orientations in a given part of the visual field is considered, the response must be considered as a spatio-temporal wave. We propose a neural field model consisting of FitzHugh-Nagumo neurons, that generates such a wave. Reflecting the dynamics of a single FitzHugh-Nagumo neuron, the neural field also exhibits excitatory and oscillatory regimes as an offset parameter is increased. We consider the question of the manner in which the input orientation is coded in the response of the neural field and discovered that two different codes − Amplitude Modulation and Phase Modulation − are present. Whereas for smaller offset values, when the model is in excitatory regime the orientation is coded in terms of amplitude, for larger offset values when the model is in the oscillatory regime, the orientation is coded in terms of phase.