Scintera Networks, Inc.

Sri International

University of Southern California

Google Inc.

Inphi Corporation

Waymo LLC

Maxim Integrated Products, Inc.

Other

Honeywell International Inc.

Qian Yu

Abhijit G Shanbhag

Debanjan Mukherjee

Abhijit M Phanse

Jishnu Bhattacharjee

Venugopal Balasubramonian

Edem Ibragimov

Yan Wang

Fabian Giroud

Harpreet Singh Sawhney

Ajay Divakaran

Prashant Choudhary

Amir Tamrakar

Rajeev Krishnamoorthy

Hui Ting Cheng

Alan Eli Willner

Martin Christian Stumpe

Omar Javed

Liron Yatziv

Jingen Liu

Olivier Charlon

Vinay Damodar Shet

Gerard Medioni

Lianshan Yan

Sacha Christophe Arnoud

Isaac D Cohen

Dar-Shyang Lee

Shayan Farahvash

Congcong Li

Christian Szegedy

Junhua Mao

Oded Fuhrmann

Jayakrishnan Kumar Eledath

Zhongqi Pan

Madabusi Govindarajan

Jiejie Zhu

Saad Ali

Adric Quade Broadwell

Weiyu Zhang

Yunqian Ma

Dumitru Erhan

Frederic Roger

Sanggeon Lee

Armando C Cova

Asaf B Sahin

Thang Ba Dinh

Frédéric Roger

Venu Balasubramonian

Jishnu Bhatacharjee

Ting Luo

Yeqing Li

Changyuan Yu

Sacha Christopher Arnoud

Some aspects of the subject matter disclosed herein include a system implemented on one or more data processing apparatuses. The system can include an interface configured to obtain, from one or more sensor subsystems, sensor data describing an environment of a vehicle, and to generate, using the sensor data, (i) one or more first neural network inputs representing sensor measurements for a particular object in the environment and (ii) a second neural network input representing sensor measurements for at least a portion of the environment that encompasses the particular object and additional portions of the environment that are not represented by the one or more first neural network inputs; and a convolutional neural network configured to process the second neural network input to generate an output, the output including a plurality of feature vectors that each correspond to a different one a plurality of regions of the environment.

Object classification using extra-regional context

Aspects of the disclosure provide for a method for updating business information. A business at a business location and a reference image for the business at the business location may be selected for an update. From a number of more recent images, a comparison image may be selected based on a likelihood that the comparison image captures the business location. Text and visual features reference image may be compared with text and visual features of the comparison image to determine a text similarity score and a features similarity score. A confidence level indicating whether the business in the reference image is in the comparison image may then be determined using the text similarity score and the feature similarity score. According to the confidence level, the business information of the business may be updated.

Business change detection from street level imagery

A system for object detection and tracking includes technologies to, among other things, detect and track moving objects, such as pedestrians and/or vehicles, in a real-world environment, handle static and dynamic occlusions, and continue tracking moving objects across the fields of view of multiple different cameras.

Real-time detection, tracking and occlusion reasoning

A complex video event classification, search and retrieval system can generate a semantic representation of a video or of segments within the video, based on one or more complex events that are depicted in the video, without the need for manual tagging. The system can use the semantic representations to, among other things, provide enhanced video search and retrieval capabilities.

Classification, search and retrieval of complex video events

Systems and methods of identifying entities are disclosed. In particular, one or more images that depict an entity can be identified from a plurality of images. One or more candidate entity profiles can be determined from an entity directory based at least in part on the one or more images that depict the entity. The one or more images that depict the entity and the one or more candidate entity profiles can be provided as input to a machine learning model. One or more outputs of the machine learning model can be generated. Each output can include a match score associated with an image that depicts the entity and at least one candidate entity profile. The entity directory can be updated based at least in part on the one or more generated outputs of the machine learning model.

Systems and methods for identifying entities directly from imagery

A food recognition assistant system includes technologies to recognize foods and combinations of foods depicted in a digital picture of food. Some embodiments include technologies to estimate portion size and calories, and to estimate nutritional value of the foods. In some embodiments, data identifying recognized foods and related information are generated in an automated fashion without relying on human assistance to identify the foods. In some embodiments, the system includes technologies for achieving automatic food detection and recognition in a real-life setting with a cluttered background, without the images being taken in a controlled lab setting, and without requiring additional user input (such as user-defined bounding boxes). Some embodiments of the system include technologies for personalizing the food classification based on user-specific habits, location and/or other criteria.

Automated food recognition and nutritional estimation with a personal mobile electronic device

Real-time object detection, tracking and occlusion reasoning

A system includes a crest-factor reduction circuit, a signal analyzer, and a pre-distortion circuit. The crest-factor reduction circuit reduces a crest factor of a baseband signal and generates a feedforward signal. The signal analyzer generates parameters based on the feedforward signal and an output signal from a power amplifier. The pre-distortion circuit generates a pre-distorted baseband signal based on the parameters for input to the power amplifier.

Adaptively controlled pre-distortion circuits for RF power amplifiers

Aspects of the present disclosure relate to a method includes training a deep neural network using training images and data identifying one or more business storefront locations in the training images. The deep neural network outputs tight bounding boxes on each image. At the deep neural network, a first image may be received. The first image may be evaluated using the deep neural network. Bounding boxes may then be generated identifying business storefront locations in the first image.

Business discovery from imagery

A transceiver coupled to an antenna includes: (a) a multi-port filter having a bidirectional port coupled to the antenna, at least one input port and at least one output port; (b) a transmit datapath receiving a transmission signal and providing the transmission signal for transmission by the antenna through the multi-port filter, the transmit datapath being coupled to the input port of the multi-port filter; (c) a receive datapath receiving a reception signal from the antenna, the receive datapath being coupled to the output port of the multi-port filter; (d) a band-pass filter coupled to the antenna for receiving a sampled signal that includes intermodulation components between two or more of an external signal, the transmission signal and the reception signal; and (e) a monitoring and cancellation circuit receiving the transmission signal, the reception signal and the sampled signal to cancel the intermodulation components.

In-service monitoring and cancellation of passive intermodulation interferences

A system includes an interleaved analog-to-digital converter (ADC) comprising a plurality of sub-ADCs, where each of the plurality of sub-ADCs has an adjustable timing. The system includes a data analyzer that analyzes an output of the interleaved ADC, that estimates timing mismatches of the plurality of sub-ADCs, and that corrects the timing mismatches by adjusting the adjustable timing of one or more of the plurality of sub-ADCs based on the estimated timing mismatches.

Adaptive correction of interleaving errors in time-interleaved analog-to-digital converters

Methods and apparatus for blind detection and correction of interleaving errors using all-digital processing of data output by multiple sub-ADCs of a time-interleaved ADC are disclosed. The methods and apparatus detect and correct frequency-dependent timing and gain mismatches of the sub-ADCs of the time-interleaved ADC using a novel multiple-input multiple-output (MIMO) filter structure. Additionally, a novel adaptive sampling frequency hopping scheme is proposed to achieve uninterrupted data conversion by the time-interleaved ADC during transient time of ADC clock frequency switching.

Interleaving error correction and adaptive sample frequency hopping for time-interleaved analog-to-digital converters

Growing community of inventors

Santa Clara, CA, United States of America

Qian Yu