Sensormatic Electronics Corporation

Rethink Robotics Gmbh

Intellivid Corporation

Microsoft Technology Licensing, LLC

Mftb Holdco, Inc.

Other

Johnson Controls Tyco IP Holdings LLP

Zillow, Inc.

Johnson Controls Gmbh

Johnson Controls US Holdings LLC

Christopher J. Buehler

Christopher J Buehler

Michael Siracusa

Matthew Murray Williamson

Rodney Brooks

Ivaylo Boyadzhiev

Lei Zhang

Yuguang Li

Noelle Dye

Bruce M Blumberg

Yuxiao Hu

Lambert E Wixson

Cornelia Carapcea

Gerry Ens

Michael Sussman

Pierre Moulon

Natan Linder

Matthew DiCicco

Joseph M Romano

William James Dally

Sing Bing Kang

Neel Suresh Joshi

Ian Christopher Westmacott

Patrick Sobalvarro

Michael E Caine

Yandong Guo

Howard I Cannon

Neil Brock

Elaine Chen

Scott W Rixner

Will Adrian Hutchcroft

Matthew A Gruenke

Sean P Cier

Jeffrey P Grossman

William A Goodwin

Matthew M Williamson

Joshua Weisberg

Patrick A Deegan

Anna Roth

Paula Long

Wei-Sheng Lai

Albert Huang

Yujia Huang

Jehanbux Edulbehram

Elaine Y Chen

Howard J Cannon

Techniques are described for using computing devices to perform automated operations for using generated building information to further generate and present visual data enhancements on images that are captured by and displayed on a mobile device in the building (e.g., concurrent with capture of the images by a camera of the mobile device, such as in a real-time or near-real-time manner with respect to image capture), such as using descriptive information about a building that is generated from analysis of acquired building images and optionally other building information (e.g., floor plans), and such as to improve navigation of the building (e.g., for autonomous vehicles) and provide other functionality as the mobile device moves through the building. The descriptive building information may include structural elements and other objects identified in the building and other determined attributes of the building, and automatically generated textual descriptions of the objects and other attributes.

Automated tracking of in-building device location and provision of associated location-related functionality

Techniques are described for automated operations related to analyzing visual data from images captured in rooms of a building and optionally additional captured data about the rooms to assess room layout and other usability information for the building's rooms and optionally for the overall building, and to subsequently using the assessed usability information in one or more further automated manners, such as to improve navigation of the building. The automated operations may include identifying one or more objects in each of the rooms to assess and one or more target attributes of each object, and generating and providing instructions to cause the capture of images and optionally additional data about those objects and target attributes for further automated analysis, such as instructions provided to an automated image acquisition device in the room and/or an associated user. Once received, the captured images and other data is then analyzed accordingly.

Automated direction of capturing in-room information for use in usability assessment of buildings

A method includes obtaining training data for a classifier, the training data comprises one or more target classes, obtaining candidate background classes, selecting negative classes from the candidate background classes, wherein the negative classes exclude candidate background classes that are close to the target classes, wherein the negative classes exclude candidate background classes that are very different from the target classes, and wherein the negative classes include candidate background classes that are similar to the target classes, and training the classifier on a combined set of the selected negative classes and target classes.

Adding negative classes for training classifier

Techniques are described for computing devices to perform automated operations related to using images acquired in a building as part of generating a floor plan for the building, in some cases without using depth information from depth-sensing equipment about distances from the images' acquisition locations to objects in the surrounding building, and for subsequent use in further automated manners, such as controlling navigation of mobile devices and/or for display to end users in a corresponding GUI (graphical user interface). In some cases, the MIGM system interacts with an MIGM system operator user, such as by displaying a GUI showing information related to the images and/or a floor plan being generated, and by receiving and using input submitted by the user via the GUI to assist with the generating of the floor plan, such as to specify interconnections between particular rooms via particular inter-room wall openings of the rooms.

Automated tools for generating building mapping information

Techniques are described for automated operations to analyze visual data combined from multiple images captured in a room to determine the room shape, such as by iteratively refining alignment of the multiple images' visual data into a common coordinate system until alignment differences satisfy one or more defined criteria, and for subsequently using the determined room shape information in further automated manners. The images may be panorama images in an equirectangular or other spherical format, and determined room shapes for one or more rooms of a building may be fully closed three-dimensional shapes and used to improve navigation of the building (e.g., as part of a generated building floor plan)&#x2014;the automated room shape determination may be further performed without having or using information from any distance-measuring devices about distances from an image's acquisition location to walls or other objects in the surrounding room.

Automated room shape determination using visual data of multiple captured in-room images

A video monitoring and analysis system detect subjects when they are entering and/or exiting from a room. The system enables a user to define a portal, such as doorway of the room. The system then monitors the movement of foreground objects in the room. Objects that appear only in the portal are classified as passing by the portal, e.g., doorway. Objects that initially appear in the portal and then are detected moving within the room are classified as having entered the room. Objects that are in the room and then disappear within the portal are classified as having exited the room. The system further has provisions for generating real-time alerts and performing forensic searches.

Method and system for monitoring portal to detect entry and exit

Methods, systems, and computer programs are presented for training a deep neural network (DNN). One method includes an operation for training a predecessor network defined for image recognition of items, where parameters of a predecessor classifier are initialized with random numbers sampled from a predetermined distribution, and the predecessor classifier utilizes an image-classification probability function without bias. The method further includes an operation for training a successor network defined for image recognition of items in a plurality of classes, where parameters of a successor classifier are initialized with parameters learned from the predecessor network, and the successor classifier utilizes the image-classification probability function without bias. Further, the method includes operations for receiving an image for recognition, and recognizing the image utilizing the successor classifier.

Fast deep neural network training

Hyperlapse results are generated from wide-angled, panoramic video. A set of wide-angled, panoramic video data is obtained. Video stabilization is performed on the obtained set of wide-angled, panoramic video data. Without user intervention, a smoothed camera path is automatically determined using at least one region of interest that is determined using saliency detection and semantically segmented frames of stabilized video data resulting from the video stabilization. A set of frames is determined to vary the velocity of wide-angled, panoramic rendered display of the hyperlapse results.

Controlling generation of hyperlapse from wide-angled, panoramic videos

An integrated surveillance system combining video surveillance and data from other sensor-based security networks is used to identify activities that may require attention.

Object tracking and alerts

Via intuitive interactions with a user, robots may be trained to perform tasks such as visually detecting and identifying physical objects and/or manipulating objects. In some embodiments, training is facilitated by the robot's simulation of task-execution using augmented-reality techniques.

Vision-guided robots and methods of training them

Growing community of inventors

Cambridge, MA, United States of America

Christopher J Buehler