Pattern Recognition in a Bucket
نویسندگان
چکیده
This paper demonstrates that the waves produced on the surface of water can be used as the medium for a “Liquid State Machine” that pre-processes inputs so allowing a simple perceptron to solve the XOR problem and undertake speech recognition. Interference between waves allows non-linear parallel computation upon simultaneous sensory inputs. Temporal patterns of stimulation are converted to spatial patterns of water waves upon which a linear discrimination can be made. Whereas Wolfgang Maass’ Liquid State Machine requires fine tuning of the spiking neural network parameters, water has inherent self-organising properties such as strong local interactions, time-dependent spread of activation to distant areas, inherent stability to a wide variety of inputs, and high complexity. Water achieves this “for free”, and does so without the time-consuming computation required by realistic neural models. An analogy is made between water molecules and neurons in a recurrent neural network.
منابع مشابه
A robust gross-to-fine pattern recognition system
This paper presents a model-based vision recognition engine for planar contours that are scale invariant of known models. Features are obtained by using a constant-curvature criterion and used to carry out efficient coarse-to-fine recognition. A robust shape matching is proposed for comparing contour fragments from scenes with partial occluding. In order to carry out an early pruning of a large...
متن کاملThe eccentric connectivity index of bucket recursive trees
If $G$ is a connected graph with vertex set $V$, then the eccentric connectivity index of $G$, $xi^c(G)$, is defined as $sum_{vin V(G)}deg(v)ecc(v)$ where $deg(v)$ is the degree of a vertex $v$ and $ecc(v)$ is its eccentricity. In this paper we show some convergence in probability and an asymptotic normality based on this index in random bucket recursive trees.
متن کاملOn high resolution positioning of straight patterns via multiscale matched filtering of the Hough transform
We present a new algorithm for sub-bucket resolution segment positioning based on multiscale matched ltering of the Hough transform. It exhibits satisfactory performance even in the presence of noise and does not depend on feature width, thus being attractive for many image processing applications.
متن کاملLocal gradient pattern - A novel feature representation for facial expression recognition
Many researchers adopt Local Binary Pattern for pattern analysis. However, the long histogram created by Local Binary Pattern is not suitable for large-scale facial database. This paper presents a simple facial pattern descriptor for facial expression recognition. Local pattern is computed based on local gradient flow from one side to another side through the center pixel in a 3x3 pixels region...
متن کاملComplicated Bunch Pattern in Pep-ii*
PEP-II, the asymmetric B-Factory at SLAC, has delivered a luminosity of 3.1 · 10 cm s. This was achieved with 692 bunches in a basic "by-3" bunch pattern. Many gaps in the bunch train were necessary to suppress a blow-up of the positron beam due to an electron cloud instability (ECI). The actual pattern had 10 bunches in a row (every third bucket), then 6 bunches missing. This 48-bucket long pa...
متن کاملThe Behavior of Suction Bucket under Cyclic Loading in Saturated Lands
Numerical modeling was carried out to investigate the responses of suction bucket foundations under vertical cyclic loading. Numerical results were compared with centrifuge experiment results. Before the capacity of the model suction bucket to resist cyclic loading could be investigated, the static pullout capacities had to be determined. The buckets with different aspect ratio are investigated...
متن کامل