Massachusetts Institute of Technology

Mitsubishi Electric Research Labs.

Charmed.com, Inc.

Alex Paul Pentland

Peter Krafft

Pattie E Maes

Thad E Starner

Esteban Moro

Daniel Calacci

Yaniv Altshuler

Abhimanyu Dubey

Dhaval Adjodah

Yan Leng

Thad Eugene Starner

Nadav Aharony

Yuri A Ivanov

Baback Moghaddam

Bruce M Blumberg

Kenneth B Russell

Brian L Jordan

Trevor J Darrell

Nathan Norfleet Eagle

Christopher R Wren

Alexander Lightman

Russell Eugene Hoffman

Jackson Jarrell Pair

Ali Azarbayejani

Wei Pan

Bradley J Rhodes

Matthew Turk

Eero P Simoncelli

Thomas P Stephenson

Stanley E Sclaroff

Michael P Johnson

Tinsley Galyean

In some implementations of this invention, the performance of a network of reinforcement learning agents is maximized by optimizing the communication topology between the agents for the communication of gradients, weights or rewards. For instance, a sparse Erdos-Renyi network may be employed, and network density may be selected in such a way as to maximize reachability and to minimize homogeneity. In some cases, a sparse network topology is employed for massively distributed learning, such as across entire fleets of autonomous vehicles or mobile phones that learn from each other instead of requiring a master to coordinate learning.

Methods and apparatus for communication network

In exemplary implementations of this invention, mobile application (app) installations by users of one or more networks are predicted. Using network data gathered by smartphones, multiple 'candidate' graphs (including a call log graph) are calculated. The 'candidate' graphs are weighted by an optimization vector and then summed to calculate a composite graph. The composite graph is used to predict the conditional probabilities that the respective users will install an app, depending in part on whether the user's neighbors have previously installed the app. Exogenous factors, such as the app's quality, may be taken into account by creating a virtual candidate graph. The conditional probabilities may be used to select a subset of the users. Signals may be sent to the subset of users, including to recommend an app. Also, the probability of successful &#x201c;trend ignition&#x201d; may be predicted from network data.

Methods and apparatus for prediction and modification of behavior in networks

In exemplary implementations of this invention, one or more computer processors receive electronic data indicative of, or compute (i) at least three different topologies of a network and (ii) a level of network performance of a task for each of the different topologies, respectively. The processors also calculate (i) a cascade probability for each of the different topologies, respectively, (ii) a curve indicative of correlation between the cascade probabilities and levels of network performance, and (iii) an optimal cascade probability which optimizes the level of network performance. A topological change in the network is produced (or its likelihood is increased). The topological change makes or would make the cascade probability closer to the optimal cascade probability. The processors output control signals (i) to make the topological change or (ii) to communicate an incentive for the topological change to an electronic node device in the network.

Methods and apparatus for tuning a network for optimal performance

A method of altering a social signaling characteristic of a speech signal. A statistically large number of speech samples created by different speakers in different tones of voice are evaluated to determine one or more relationships that exist between a selected social signaling characteristic and one or more measurable parameters of the speech samples. An input audio voice signal is then processed in accordance with these relationships to modify one or more of controllable parameters of input audio voice signal to produce a modified output audio voice signal in which said selected social signaling characteristic is modified. In a specific illustrative embodiment, a two-level hidden Markov model is used to identify voiced and unvoiced speech segments and selected controllable characteristics of these speech segments are modified to alter the desired social signaling characteristic.

Modification of voice waveforms to change social signaling

Portable communication devices, such as Bluetooth enabled cellular phones, communicate with and identify like devices that are nearby, and send notification messages to a remote server. When a notification message is received at the server identifying two devices that have come within range of one another, the server compares the profile data associated with each of the two identified devices and facilitates communications between the devices when appropriate.

Combined short range radio network and cellular telephone network for interpersonal communications

The location of an object in a physical scene is identified with a stereo camera. A virtual surface is identified in the physical scene, and an approximate disparity set is constructed for the virtual surface. A main and a reference image of the scene are acquired by the stereo camera. The reference image is warped according to the disparity set, and subtracted from the main image to determine a set of depth residual values. Pixels having a substantially non-zero residual are identified with a surface of the object not coincident with the virtual surface.

Computer vision depth segmentation using virtual surface

Techniques and approaches that facilitate acquisition, transmission or retrieval of data for wearable devices are disclosed. These wearable devices are electronic devices, such as mobile computing devices or wireless communication devices, and are often small in scale and very portable. Wearable devices are able to communicate with one another to exchange information. Wearable devices are also able to exchange information with a portal server. Personal portals can also be provided for users of the wearable devices so that they can easily access information gather by their wearable device and subsequently transmitted to their personal portal.

Method and system for data transmission between wearable devices or from wearable devices to portal

Documents stored in a database are searched for relevance to contextual information, instead of (or in addition to) similar text. Each stored document is indexed in term of meta-information specifying contextual information about the document. Current contextual information is acquired, either from the user or the current computational or physical environment, and this &ldquo;meta-information&rdquo; is used as the basis for identifying stored documents of possible relevance.

Method and apparatus for automated, context-dependent retrieval of information

Methods and apparatus for detecting instances of a selected object or object feature in a digitally represented scene utilize analysis of probability densities to determine whether an input image (or portion thereof) represents such an instance. The invention filters images of objects that, although in some ways similar to the object under study, fail to qualify as typical instances of that object. The invention is useful in the detection and recognition of virtually any multifeatured entity such as human faces, features thereof (e.g., eyes), as well as non-rigid and articulated objects such as human hands.

Detection, recognition and coding of complex objects using probabilistic

In a morphing system for creating intermediate images that, viewed serially, make an object in a source image appear to metamorphose into a different object in a target image, correspondences between feature points in the source and target images are determined by simulating the modes of motion of respective elastic sheets in which are embedded nodal points that correspond to respective feature points in the images. For each feature point, a generalized-feature vector is determined whose components represent the associated nodal point's participations in the various motion modes. Correspondences between feature points in the source and target images are determined in accordance with the closeness of the points' generalized feature vectors. Correspondences thus determined can additionally be used for alignment and object-recognition purposes.

Finite-element method for image alignment and morphing

An electronic system for integrating a user's moving image into a computer-generated pictorial environment, and allowing the user to interact with the environment and objects therein. The environment is rendered on a screen display in a manner that simulates three dimensions, including respect for occlusion relationships among the user and objects in the environment. The resulting image, which changes as the user gestures and interacts with the environment, is projected onto a large screen that the user faces. The environment includes a series of mobile entities, or 'agents,' which respond to one or more of the user's changing position, gestures and voice, exhibiting context-dependent behavior that appears semi-intelligent.

Method and system for facilitating wireless, full-body, real-time user

A method and apparatus for generating three-dimensional, textured computer models from a series of video images of an object is disclosed. The invention operates by tracking a selected group of object features through a series of image frames and, based on changes in their relative positions, estimates parameters specifying camera focal length, translation and rotation, and the positions of the tracked features in the camera reference frame. After segmentation of the images into two-dimensional bounded regions that each correspond to a discrete surface component of the actual object, the texture contained in the various video frames is applied to these regions to produce a final three-dimensional model that is both geometrically and photometrically specified.

Method and apparatus for three-dimensional, textured models from plural

An apparatus for estimating the solution to a finite element analysis and/or regularization equation, the apparatus including means for generating a pyramid representation of input data, each level of the pyramid representation corresponding to a different set of similar eigenvectors; means for multiplying at least one of the levels of the representation by a weight derived from the eigenvalue associated with the set of eigenvectors for that level to generate a weighted pyramid representation; and means for collapsing the weighted pyramid representation to generate an estimated solution to the equation.

Method for estimating solutions to finite element equations by

A recognition system for identifying members of an audience, the system including an imaging system which generates an image of the audience; a selector module for selecting a portion of the generated image; a detection means which analyzes the selected image portion to determine whether an image of a person is present; and a recognition module responsive to the detection means for determining whether a detected image of a person identified by the detection means resembles one of a reference set of images of individuals.

Growing community of inventors

Lexington, MA, United States of America

Alex Paul Pentland