Abstract: A system and method for detecting and filtering non-violation events in a traffic light violation prediction and recording system, including at least one violation prediction image capturing device, such as a video camera, and a violation prediction unit. The prediction unit generates a prediction reflecting a probability that the vehicle will violate a red light phase of the traffic signal. A non-violation event filter determines whether the vehicle approaching the traffic signal is actually performing a non-violation action. Non-violation events may include a variety of actions performed by the vehicle, and are fully configurable to meet the needs and policies of various specific intersections and jurisdictions. When the non-violation event filter determines that the vehicle is performing a non-violation action, it causes some number of resources that may have been allocated to recording the non-violating vehicle, and/or prevents further resources from being allocated to such recording.

Abstract: A system for generating a traffic violation citation, which operates using at least one image data file including digitized video data. The image data file includes multiple images of an intersection during a red light violation. The image data file is, for example, derived from the output of one or more video cameras positioned to record respective views of an intersection being monitored for red light violations. The disclosed system provides a user with an interface including at least one image display window. The image display window is used to display portions of the image data file. Where the image data file includes multiple video clips, such clips may be displayed simultaneously in respective ones of multiple display windows. The system further includes a selection interface that permits a user to select at least one image from the image data file, for example a current image being displayed in an image display window.

Abstract: A system and method for detecting and filtering non-violation events in a traffic light violation prediction and recording system, including at least one violation prediction image capturing device, such as a video camera, and a violation prediction unit. The prediction unit generates a prediction reflecting a probability that the vehicle will violate a red light phase of the traffic signal. A non-violation event filter determines whether the vehicle approaching the traffic signal is actually performing a non-violation action. Non-violation events may include a variety of actions performed by the vehicle, and are fully configurable to meet the needs and policies of various specific intersections and jurisdictions. When the non-violation event filter determines that the vehicle is performing a non-violation action, it causes some number of resources that may have been allocated to recording the non-violating vehicle, and/or prevents further resources from being allocated to such recording.

Abstract: A system and method for detecting and filtering non-violation events in a traffic light violation prediction and recording system, including at least one violation prediction image capturing device, such as a video camera, and a violation prediction unit. The prediction unit generates a prediction reflecting a probability that the vehicle will violate a red light phase of the traffic signal. A non-violation event filter determines whether the vehicle approaching the traffic signal is actually performing a non-violation action. Non-violation events may include a variety of actions performed by the vehicle, and are fully configurable to meet the needs and policies of various specific intersections and jurisdictions. When the non-violation event filter determines that the vehicle is performing a non-violation action, resources that may have been allocated to recording the non-violating vehicle may be released, and a determination may be made that additional resources need not be allocated to such recording.

Abstract: A traffic light violation prediction and recording system, including at least one violation prediction video camera and a violation prediction unit. The violation prediction unit generates violation probability scores for vehicles approaching a traffic intersection based on attributes of those vehicles, such as position, current speed and acceleration. The violation probability scores are passed to a violation recording unit, which allocates violation recording resources, such as a violation video camera, to capture activities of one or more vehicles associated with relatively high violation probabilities. The violation recording unit determines a violation recording resource allocation schedule. If sufficient resources cannot be allocated to record all possible violations within a given time period, the violation recording unit may ignore a potential violation having a relatively low priority.

Abstract: A recognition system is disclosed, including a representation of an object in terms of its constituent parts that is translationally invariant, and which provides scale invariant recognition. The system further provides effective recognition of patterns that are partially present in the input signal, or that are partially occluded, and also provides an effective representation for sequences within the input signal. The system utilizes dynamically determined, context based expectations, for identifying individual features/parts of an object to be recognized. The system is computationally efficient, and capable of highly parallel implementation, and further includes a mechanism for improving the preprocessing of individual sections of an input pattern, either by applying one or more preprocessors selected from a set of several preprocessors, or by changing the parameters within a single preprocessor.

Abstract: A collision avoidance system and method for a traffic intersection having a traffic light, which selectively extends the duration of a red light cycle to prevent traffic from entering the intersection during a red light violation. Output from one or more violation prediction image capturing devices is used to provide images showing one or more vehicles approaching the intersection. A violation prediction unit, which is also coupled to a signal reflecting a current light phase of the first traffic signal, receives and processes vehicle location information derived from the images to generate violation predictions for the vehicles approaching the intersection. The violation prediction indicates a likelihood that an associated vehicle will violate an upcoming red light phase of the traffic signal. In response to the violation prediction, a violation predicted signal is provided to a traffic signal controlling traffic intersecting with the predicted violator.

Abstract: A system and method for integrating traffic light violation related vehicle information with court date scheduling information, which employs one or more digitized video recordings of alleged violations. The system provides end-to-end violation processing using the digitized video recordings of the event, and using vehicle database information, as may for example be provided from a department of motor vehicles (DMV) database. The system potentially also interfaces to a court date scheduling database, which includes indications of various available times at which hearings regarding traffic violations may be scheduled. A user interface may provide a display of one or more perspectives on the violation event, as well as access to the vehicle and court date scheduling database.

Abstract: A neural network including a plurality of sub-nets stored in a memory array and a method of operation. Each of the sub-nets includes a corresponding plurality of weights and is individually operable to classify an input vector. A computation unit, including a distance calculation unit and a math unit, is responsive to an input vector comprising input training features for determining a distance between the weights of each sub-net to each of the input features of the input vector and for determining whether the distance is within a particular region of influence. Also described is a parallel process for training the sub-nets.

Abstract: A neural network and method for pipeline operation within a neural network which permits rapid classification of input vectors provided thereto is disclosed. In a training mode, a plurality of training input features are presented to the neural network and distances between the plurality of training features and a plurality of prototype weight values are concurrently computed. In response to an indication of a last training epoch count values for each of the prototype weight values are stored in a memory to thereby allow the neural network to operate in a probabilistic classification mode.

Abstract: A system is disclosed for separating and identifying classes of patterns or events which are not necessarily linearly separable. The patterns are represented by an input signal S. The system comprises (1) a plurality of classification units, connected in parallel to receive the input signal S and (2) a class selection device, responsive to the output signals produced by the classification units, for producing a single output response R representing the class of each respective pattern. At least some of the pattern classification units include generalizer units having a memory for storing a number of generalizer "prototypes" and a comparator for comparing the vector location of an input pattern with each of the generalizer prototypes.

Abstract: A learning algorithm for the N-dimensional Coulomb network is disclosed which is applicable to multi-layer networks. The central concept is to define a potential energy of a collection of memory sites. Then each memory site is an attractor of other memory sites. With the proper definition of attractive and repulsive potentials between various memory sites, it is possible to minimize the energy of the collection of memories. By this method, internal representations may be "built-up" one layer at a time. Following the method of Bachmann et al. a system is considered in which memories of events have already been recorded in a layer of cells. A method is found for the consolidation of the number of memories required to correctly represent the pattern environment. This method is shown to be applicable to a supervised or unsupervised learning paradigm in which pairs of input and output patterns are presented sequentially to the network.

Abstract: A system is disclosed for separating and identifying classes of patterns or events which are not necessarily linearly separable. The patterns are represented by an input signal S. The system comprises (1) a plurality of classification units, connected in parallel to receive the input signal S, and (2) a class selection device, responsive to the output signals produced by the classification units, for producing a single output response R representing the class of each respective pattern. In a preferred embodiment each classification unit includes a pattern feature encoding device, responsive to the input signal S, for producing an intermediate signal F.sub.i representative of the features contained in the pattern, and a pattern classifier, responsive to the signal F.sub.i, for producing an output signal R.sub.i representative of the respective class of the pattern, as identified by the features encoded by the feature encoding device.