Abstract: Cameras and image processing techniques are applied to the control of HVAC systems. Occupancy is detected using head-counting or motion detection. Activities are recognized in images and image sequences by machine-recognition techniques. The nature of activities, the intensity of activities, the number of occupants and their activities, etc. are all inferred from images and image sequences and used to predict current loads and/or required control signals for regulating an HVAC system.

Abstract: A vision system that is capable of computing the crowd densities, shopper movement speed and direction, and queue lengths at different locations in a shopping center makes recommendations for least-cost path through a set of destinations. Well-known person-counting techniques may be employed. One type of output display is a map display with an overlay an optimal route through a shopping venue.

Abstract: An acoustic based pointing device and a system and method for using the pointing device to move a cursor on a display screen of a display device. The pointing device comprises an acoustic source that generates and propagates an acoustic signal that is detected by four microphones M1, M2, M3, and M4 and is then converted into corresponding periodic electrical signals S1, S2, S3, and S4. The position vector P0 of the acoustic source is calculated by solving triangulation equations that depend on phase-shift time delays between S1 and Sj for j=1, 2, and 3. A position vector PC is calculated from P0 using a scale vector that relates a change in position of the cursor to a change in position of the acoustic source. Then the cursor is moved to a position PC associated with the position vector PC.

Abstract: A system and method for cordless communication between a modem of a computer and a communication line includes a base station and at least a remote unit. A booster station can be optionally included to increase reception to the remote unit. The base station will be adapted for connection with the communication line, and will transmit via RF to the remote unit. The base station may test the communications over a plurality of available channels and select the channel with the strongest reception. The booster station may also communicate via RF with the remote unit, and may optionally plug into an electrical outlet, and communicate with the base station over common wiring when the base station is similarly attached to another outlet of the common wiring. The system may compare reception between the base station and the remote unit with the booster station and the remote unit, and may communicate with the booster station only over the electrical wires to reduce RF transmissions.

Abstract: A system and method is disclosed for taking the results of a search (i.e., electronic documents), using the TFIDF of terms contained in those search result documents to quantify how related the search result documents are to one another, and organizing the search results based on those quantified relations. Using such a system and method, the search result documents can be displayed in such a manner that the relationships between the search result documents is evident to the viewer.

Abstract: Background-foreground segmentation is performed as a maximum likelihood classification. During a training procedure, a system estimates the parameters of likelihood probability models, which are the probability of observing images assuming that the images come from the background scene. During normal operation, the likelihood probability of captured images is estimated using the background models. The background-foreground segmentation is carried out by comparing the likelihood probabilities of the test images with fixed thresholds. The probability of observing foreground objects is assumed constant, as foreground images are generally not modeled. This value, the probability threshold, preferably represents a tunable parameter of the system. Pixels with low likelihood probability of belonging to the background scene are classified as foreground, while the rest are labeled as background.

Abstract: A media player controller is disclosed that monitors non-user event and automatically controls a media player in response to predefined events. The media player controller includes an information capture device that obtains input information from the non-user. The input information is processed by the media player controller to identify one or more predefined non-user events. A number of rules can be utilized to define various non-user events, such as when the non-user is in the vicinity of the media player. Each rule contains a condition and a corresponding action-item that should be performed when the condition is satisfied.

Abstract: The position of a target lying on a plane is indicated by inputting a pointing direction which is computationally projected onto a plane. If the target is known to lie on a contour, the position is specified unambiguously. Alternatively, the position of a target can be specified by inputting its axial coordinates in successive steps. In another alternative approach, the image containing the target is translated and/or rotated and the target indicated again. The two direction projections are then used to determine the position of the target in 2-space. The direction indications may be input by a camera or other method, such as one or more radio transmitters.

Abstract: The learning rate used for updating the weights of a self-ordering feature map is determined by a process that injects some type of perturbation into the value so that it is not simply monotonically decreased with each training epoch. For example, the learning rate may be generated according to a pseudorandom process. The result is faster convergence of the synaptic weights.

Abstract: Aspects of the present invention can be configured to demultiplex an asynchronously multiplexed video signal, which comprises images from a number of different cameras. Image matching techniques are used to assign input images with states. After a period, the number of states will generally equal the number of input cameras. The states may be modeled through any number of techniques, such as histogram analysis, clustering, and hidden Markov model analysis. Input images are assigned to states, and the input images are output as being associated with the states. Zone surveillance may be performed on a series of images from one or more of the states. Any events that occur can be distinguished and reported.

Abstract: A lighting controller is disclosed that automatically adjusts a lighting device based on user activity. The disclosed lighting controller processes at least one of audio and video information to identify predefined user events that suggest that a lighting device should be activated or adjusted. In one implementation, a number of behavior rules define various user activities or events that suggest that the user would like to activate or adjust a lighting device. Each rule contains one or more conditions, and, optionally, a corresponding action-item that should be performed when the rule is satisfied to activate or adjust the lighting device. Upon detection of a predefined event, the corresponding action, if any, is performed by the lighting controller.

Abstract: A computer-based assisted parking system and method for parking a vehicle into a parking space includes at least one sensor arranged to sense a position of the vehicle to be parked in a predetermined area adjacent the vehicle; a sensing system for at least two-dimensional reconstruction of feedback received from the one sensor; a path planning system for determining whether positions of other vehicles adjacent the parking space provide sufficient clearance for the vehicle to be parked and providing a planned path indicating how the vehicle to be parked will be maneuvered; a control system for controlling maneuvering of the vehicle to be parked based on the planned path determined by the path planning system, and the control system controls vehicle steering, direction, speed, and application of brakes, so that the vehicle to be parked is maneuvered into the parking space. Alternatively, directions may be given as the vehicle is parked under the control of the driver.

Abstract: A process for adaptive bookmarking of often-visited web sites, comprising the steps of (a) optionally determining the identity of a particular user, (b) determining whether a webpage has been detected, (c) if the webpage in step (b) has been detected, determining whether the webpage has been previously visited by a particular user, (d) performing one of (i) creating an initial record of the webpage visit by the particular user if it has been determined in step (c) that the webpage has not been previously visited by the particular user, and (ii) determining whether the webpage has been previously bookmarked if it has been determined in step (c) that the webpage has been previously visited by the particular user, (e) updating a visitation count if it has been determined in step (c) that the webpage has been previously visited by the particular user, (f) determining whether the visitation count has reached a predetermined threshold; and (g) recommending the bookmarking of the address of the webpage if it determi

Abstract: A method and system for concatenating eigenvoice vector data with eigenface vector data to obtain a new basic eigenvector to more positively and accurately identify a person. More specifically, for a specific person, the data of an eigenface vector is concatenated with the data of an eigenvoice vector to form a single vector, and this single vector is compared to reference vectors also obtained by concatenting the data of eigenface vectors and eigenvoice vectors to obtain a single vector of each persons in a defined group of people to determine if a person that is a member of a defined group of people. Using this single eiginvector gives a higher success rate for identifying a person than an eigenface vector or an eigenvoice vector used separately or together as separate vectors.

Abstract: A system and method for alerting a driver of an automobile to a traffic condition. The system comprises at least one camera having a field of view and facing in the forward direction of the automobile. The camera captures images of the field of view in front of the automobile. A control unit receives images of the field of view from the camera and identifies objects therein of a predetermined type. The control unit analyzes the object images of at least one predetermined type to determine whether one or more of the identified objects present a condition that requires the driver's response. A display receives control signals from the control unit regarding the one or more of the objects that present a condition that requires the driver's response. The display displays an image of the object to the driver that is positioned and scaled so that it overlays the actual object as seen by the driver, the displayed image of the object enhancing a feature of the actual object to alert the driver.

Abstract: A method and system for providing a mood based virtual photo album which provides photos based upon a sensed mood the viewer. The method may include the steps of capturing a first image of a facial expression of a viewer by a camera, providing the image to a pattern recognition module of a processor, determine a mood of the viewer by comparing the facial expression with a plurality of previously stored images of facial expressions having an associated emotional identifier that indicates a mood of each of the plurality of previously stored images, retrieving a set of photos from storage for transmission to the viewer based on the emotional identifier associated with the determined mood, and transmitting the set of photos in the form of an electronic photo album.

Abstract: An image processing system is disclosed that provides automatic face or skin blurring for images. All faces or skin can be blurred, or specific faces can be blurred. In one aspect of the invention, a particular face is blurred on an image or on a series of images in a video. Faces are determined in an image, and face matching is performed to match a particular face to faces in the image. If a match is found, the face or a portion of the face is blurred in the image. The blurring is performed on a portion of the image containing the particular face. Blurring may be performed through a variety of techniques. In another aspect of the invention, voice processing is used as an adjunct to or in place of face analysis to determine if a face in an image or series of images should be blurred. In another aspect of the invention, all faces or human skin in an image or series of images is blurred.

Abstract: A system and method for virtual clothes modeling includes a unit s for obtaining a virtual image of a particular shopper, a scanner for scanning an identifying token which identifies predetermined information in a database associated with a first article of clothing, a retrieval unit for retrieving from the database a visual display of the predetermined information first article of clothing corresponding to the identifying token, and

Abstract: Disclosed is a method and system for controlling access to a television set using a remote control unit having a finger verification means. The method includes the steps of: placing at least one finger on the remote control unit for scanning; obtaining fingerprint data indicative of the scanned finger; comparing the fingerprint data to predetermined reference data to establish a fingerprint match; and, if a match is established, providing access to a number of channels associated with the scanned finger. The system includes a fingerprint reader for generating fingerprint data; a processor programmed to prompt the user to enter fingerprint data; and, a fingerprint recognition means for verifying that the user is an authorized user of the remote control apparatus.

Abstract: A surveillance and security system for automatic detection and warning of detected events includes a unit for observing behavior in a predetermined area under surveillance, a unit for processing an output of observed behavior from the unit for observing, and a includes a pattern recognition module for recognizing whether the observed behavior is associated with predefined suspicious behaviors. Upon recognition that the observed behavior is suspicious, security is notified about the potential need for increased surveillance, or automatically increased surveillance can be provided. The pattern recognition module may include infrared heat profiles of persons, images of actually people, sequences of people manipulating shopping bags, tearing sounds of tearing different types of packaging such as paper and plastic for retail products. The observation of motion, which is related to behavior, is compared against a database of predefined acts.