International Business Machines Corporation

Shanghai Zhaoxin Semiconductor Co., Ltd.

Texas Instruments Corporation

Chinese Academy of Sciences

New York University

Vestas Wind Systems A/s

Lyft, Inc.

Basf Se Corporation

Paypal, Inc.

Meta Platforms Technologies, LLC

Shanghai Jiaotong University

Tongji University

Woven Planet North America, Inc.

Shanghai Microport Medical (group) Co., Ltd.

Nio Co., Ltd.

Yi Zhou

Nathalie Baracaldo Angel

Ali Anwar

Ming Zhao

Wei Tang

Michael Paul Groover

Robin Han

Heiko Hary Ludwig

Ying Chu Wang

Forrest Samuel Briggs

Romain Clément

Edward Lin

Deming Gu

Yao Yao

Fengxia Wu

Jun Zhang

Jing Li

Bhubaneswar Mishra

Gert Karmisholt Andersen

Manoj Gupta

Ciprian Biris

Simone Bianco

Frank Bernhard Schmuck

Xiaobo Wu

Wei Zhang

Yu Liu

Liang Jie Xu

Douglas J Griffith

Soumitra Sarkar

Sebastian Sztuk

Cheng Yi Huang

Theodore Brian Vojnovich

Roger Lee Haskin

Shah Mohammad Rezaul Islam

Zhen Wang

Felipe Knop

Myung Mun Bae

Jian Xu

Robert J Curran

Qiyi Luo

Thomas K Clark

Ting Zhou

Yan Su

Bryant R Chen

Rujun Ji

Hao Meng

Junfei Li

Wen Ke Xue

Songhua Hu

Ge Shen

Vito Paolo Pastore

Michael F Priestley

Rui Liu

Uffe C Merrild

Salvatore Paxia

Kangcheng Xu

Runhua Xu

Vajira Suminda Ganepola

Haiyong Huang

Wenxing Li

Jiakuan Hu

Deirdre S Longo

Jenifer Schlotfeldt

Kirk Erik Burgess

Yin Qian

Sudhir Gurunandan Rao

Daniel Tsun Kao

Rajesh S Agarwalla

Heng Que

JinYu Lou

Gang Xu

Guoqiang Xie

Sonya Rongsheng Liang

Thirumale Niranjan

Srikanth Ramamurthy

Chunhui Liu

Lu Xiong

Michael Priestly

Chaoou Huang

Jianxin Chen

Zhuoping Yu

Tianyu Liu

Zhipeng Qin

Ofer H Gill

Sumanthkumar Sukumar

Biao Wu

Xichan Zhu

Hao Yuan

Syed Amer Zawad

Shashank Rajamoni

Qingsheng Tao

Zheng Wu

Alyson-Kathleen Riley

Chang Liang Fan

Guo Liang Yuan

Yabing Tang

Zhong Yi Zhou

Xiaoqiao Chen

Chen Sun

Honghe Tan

A mid-infrared semiconductor saturable absorber mirror based on InAs/GaSb superlattice comprises a GaSb substrate with an anti-reflection film coated on the lower surface of the GaSb substrate; InAs/GaSb superlattice which has a specific structure and thickness and is arranged on the GaSb substrate; and Bragg reflection film which is arranged on the InAs/GaSb superlattice, wherein Bragg reflection film is composed of multiple pairs of ZnS and YbFfilm layers with a thickness of ¼ wavelength, and the YbFfilm layer is connected with the InAs/GaSb superlattice. The device not only has a mid-infrared working range with a broadband operation bandwidth, but also has the advantages of designable parameters, outstanding robustness, high damage threshold and the like, and sets a foundation for the development of mid-infrared ultrafast mode-locked lasers.

Mid-infrared semiconductor saturable absorber mirror based on INAS/GASB superlattice and preparation method thereof

Disclosed are a fault diagnosis circuit, method and apparatus, and computer readable storage medium. The fault diagnosis circuit includes a safety protection circuit electrically connected to a protected circuit and a diagnosis module electrically connected to the safety protection circuit. The safety protection circuit is configured to perform check operation on stored data in the protected circuit to obtain first check data, perform error injection on second check data corresponding to the stored data, and generate a first check result signal based on the first check data and the second check data after the error injection. The diagnosis module is configured to diagnose faults in the safety protection circuit based on the first check result signal. From embodiments of this disclosure, the safety protection circuit is diagnosed faults through a simple hardware structure, to eliminate the faults based on a diagnosis result, thereby ensuring normal operation of a chip.

Circuit, method, and apparatus for diagnosing a fault, and computer readable storage medium

The present invention relates to a micro-nano structure sensitive to a laser beam in a specific direction, including a substrate, wherein an insulating layer is fixedly disposed on the substrate, the insulating layer is provided with two silicon nanowires parallel to each other and having the same shape and size, lead-out nanowires are arranged at both ends of each of the silicon nanowires and are connected with a potentiometer, and a near-field coupling effect occurs between the silicon nanowires and the substrate when laser light irradiates the silicon nanowires, and one silicon nanowire closer to a laser light source is completely suppressed and the other silicon nanowire farther away from the laser light source maintains brightness. The present invention enables precise detection of a laser signal at a specific angle and non-contact signal transmission in a specific direction. The present invention relates to a micro nano structure sensitive to a laser beam in a specific direction. The micro nano structure comprises a substrate, wherein an insulating layer is fixedly arranged on the substrate; and two silicon wires, which are parallel to each other and are in the same shape and size, are arranged on the insulating layer, and wires are led out from two ends of each silicon wire and are connected to potential meters. When laser light irradiates the silicon wires, a near field coupling effect is generated between the silicon wires and the substrate; and one silicon wire close to a laser light source is completely inhibited, and the other silicon wire far away from the laser light source maintains the brightness thereof. By means of present invention, a laser signal at a certain specific angle can be accurately detected, and non contact signal transmission can be carried out in a specific direction.

Micro-nano structure sensitive to laser beam in specific direction

A medical instrument () includes a stent () and a constraining structure (). The constraining structure () constrains, and thereby prevents tangling of, a trailing section of the stent (), ensuring successful release of the stent (). Specifically, the stent () has a first proximal end and an opposing first distal end. The first distal end is configured with an expanded configuration and a collapsed configuration. The trailing section at the first distal end includes a number of protrusions (). The constraining structure () includes a body () and a clamping mechanism (). The clamping mechanism () is disposed at a distal end of the body (), and the body () extends from the first proximal end to the first distal end. The clamping mechanism () limits relative movement and thus prevents tangling of the protrusions () by passing through all or some of the protrusions (). Additionally, it allows the stent () to be less deformed within a delivery sheath and thereby enables it to successfully expand when pushed out of the sheath.

Medical instrument

A plurality of public encryption keys are distributed to a plurality of participants in a federated learning system, and a first plurality of responses is received from the plurality of participants, where each respective response of the first plurality of responses was generated based on training data local to a respective participant of the plurality of participants and is encrypted using a respective public encryption key of the plurality of public encryption keys. A first aggregation vector is generated based on the first plurality of responses, and a first private encryption key is retrieved using the first aggregation vector. An aggregated model is then generated based on the first private encryption key and the first plurality of responses.

Privacy-preserving federated learning

A left atrial appendage (LAA) occluder and an LAA occlusion system are disclosed. The LAA occluder enables an increased surgical success rate by allowing repeated good positioning, retrieval and release of the occluder and avoiding the problem of entanglement of the occluding stent during surgery. The LAA occluder includes an occluding stent, a proximal securing member a distal securing member. The occluding stent includes an occlusion structure and a traction structure. The delivery device includes a hollow push tube and a control shaft. The distal securing member can be driven by the control shaft to move toward the proximal securing member until a first distal end of the occlusion structure is in a collapsed configuration, where the distal securing member is blocked from further movement, avoiding the problem of entanglement arising from inward rolling of the occluding stent in the vent of human error.

Left atrial appendage occluder and left atrial appendage occlusion system

A whole vehicle in-loop test system of an intelligent automobile, including: a rotary drum platform, used for simulating a longitudinal movement of a test vehicle; an environment perception platform, disposed at a front end of the rotary drum platform, used for simulating a transverse movement of a test vehicle, and including a support base at a lower portion and a vehicle placing platform at an upper portion, where a transverse drive wheel is mounted at the lower portion of the support base, and the transverse drive wheel moves along a transverse track disposed on the ground; and a target vehicle simulation unit, disposed at a periphery of the environment perception platform, and configured to simulate a target vehicle.

Whole vehicle in-loop test system of intelligent automobile

An indication of availability over time and resource usage is maintained for each computing device of a plurality of computing devices. An optimal combination of a subset of the plurality of computing devices is determined for each round of one or more rounds of training based on the availability over time and the resource usage for each computing device. A global model is generated utilizing the one or more optimal combinations of the plurality of computing devices and a query is performed utilizing the global model.

Scheduled federated learning for enhanced search

A method, a computer system, and a computer program product are provided for federated learning. An aggregator may receive cluster information from distributed computing devices. The cluster information may relate to identified clusters in sample data of the distributed computing devices. The cluster information may include centroid information per cluster. The aggregator may include a processor. The aggregator may integrate the cluster information to define data classes for machine learning classification. The integrating may include computing a respective distance between centroids of the clusters in order to determine a total number of the data classes. The aggregator may send a deep learning model that includes an output layer that has a total number of nodes equal to the total number of the data classes. The deep learning model may be for the distributed computing devices to perform machine learning classification in federated learning.

Semantic learning in a federated learning system

An image capture system is configured to align a field of view of the image capture component with a field of view of a user of the system. In some cases, the image capture system may adjust the field of view of the image data based at least in part on orientation and position data associated with the capture device.

System for determining an expected field of view

One embodiment provides a method for federated learning across a plurality of data parties, comprising assigning each data party with a corresponding namespace in an object store, assigning a shared namespace in the object store, and triggering a round of federated learning by issuing a customized learning request to at least one data party. Each customized learning request issued to a data party triggers the data party to locally train a model based on training data owned by the data party and model parameters stored in the shared namespace, and upload a local model resulting from the local training to a corresponding namespace in the object store the data party is assigned with. The method further comprises retrieving, from the object store, local models uploaded to the object store during the round of federated learning, and aggregating the local models to obtain a shared model.

Parameter sharing in federated learning

Embodiments of the present disclosure disclose a circuit detection method and a data detection circuit. The circuit detection method comprises: if a current time point reaches a preset detection time period, based on a data storage address of a detected module, reading a data to be detected corresponding to the detected module from a storage area corresponding to the data storage address; using a preset calculation method corresponding to the detected module to perform a calculation on the data to be detected to obtain a first calculation result; based on the first calculation result and a preset calculation result corresponding to the data storage address, determining a fault state of the detected module. The embodiments of the present disclosure can detect the storage circuit in a timely and accurate manner without data verification by adding hardware, thereby saving space occupied by the system and reducing power consumption.

Circuit detection method and data detection circuit

Systems, computer-implemented methods, and computer program products that can facilitate detection of an adversarial backdoor attack on a trained model at inference time are provided. According to an embodiment, a system can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can comprise a log component that records predictions and corresponding activation values generated by a trained model based on inference requests. The computer executable components can further comprise an analysis component that employs a model at an inference time to detect a backdoor trigger request based on the predictions and the corresponding activation values. In some embodiments, the log component records the predictions and the corresponding activation values from one or more layers of the trained model.

Detection of an adversarial backdoor attack on a trained model at inference time

Systems and techniques that facilitate universal and efficient privacy-preserving vertical federated learning are provided. In various embodiments, a key distribution component can distribute respective feature-dimension public keys and respective sample-dimension public keys to respective participants in a vertical federated learning framework governed by a coordinator, wherein the respective participants can send to the coordinator respective local model updates encrypted by the respective feature-dimension public keys and respective local datasets encrypted by the respective sample-dimension public keys. In various embodiments, an inference prevention component can verify a participant-related weight vector generated by the coordinator, based on which the key distribution component can distribute to the coordinator a functional feature-dimension secret key that can aggregate the encrypted respective local model updates into a sample-related weight vector. In various embodiments, the inference prevention component can verify the sample-related weight vector, based on which the key distribution component can distribute to the coordinator a functional sample-dimension secret key that can aggregate the encrypted respective local datasets into an update value for a global model.

Efficient private vertical federated learning

In one embodiment, a method includes accessing first image data generated by a first image sensor having a first filter array that has a first filter pattern. The first filter pattern includes a number of first filter types. The method also includes accessing second image data generated by a second image sensor having a second filter array that has a second filter pattern different from the first filter pattern. The second filter pattern includes a number of second filter types, the number of second filter types and the number of first filter types have at least one filter type in common. The method also includes determining a correspondence between one or more first pixels of the first image data and one or more second pixels of the second image data based on a portion of the first image data associated with the filter type in common.

Multichannel, multi-polarization imaging for improved perception

Growing community of inventors

Shanghai, China

Yi Zhou