Qfeeltech (Beijing) Co., Ltd.

Boe Technology Group Co., Ltd.

Beijing Boe Optoelectronics Technology Co., Ltd.

Baidu USA LLC

Baidu Online Network Technology (Beijing) Co., Ltd.

Beijing Baidu Netcom Science Technology Co., Ltd

University of Science and Technology Beijing

Baidu.com Times Technology (Beijing) Co. Ltd.

Microsoft Technology Licensing, LLC

Tsinghua University

Xiaomi Inc.

Chinese Academy of Medical Sciences

Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd.

Metabolic Explorer

Beijing Kansenbio Technology Ltd.

Zhen Chen

Yiming Zhang

Lei Wang

Ming Yue

Weiyao Zhu

Zhiyong Song

Shuailing Li

Liguang Deng

Manjiang Zhang

Xin Wu

Quanhua He

Davy Huang

Dehua Liu

Xing Li

Haofeng Kou

Debin Kong

Ming Chi

Huimeng Zheng

Huizhong An

Chenhao Wu

Ruichang Cheng

Shanyu Wang

Xue Dong

Haisheng Wang

Yingming Liu

Shengji Yang

Hao Zhang

Wei Li

Lingyun Shi

Jianguo Wang

Hong-Jiang Zhang

Jun Cheng

Hua Li

Chen Wang

Wei Lu

Lei Zhao

Mingjing Li

Philippe Soucaille

Guangquan Wang

Ming Zhai

Lin Liu

Yan Jiang

Jianjun Zhang

Yi Zhou

Hyungkyu Kim

Long Zhang

Wanda Dischert

Wen-Yin Liu

Kun Huang

Yang Pei

Xiaohua Cheng

Xiaohua Liu

Ge Song

Quan Liu

Yunfeng Liu

Shouke Qin

You Han

Wenchao Liu

Yuling Liu

Moyu Zhu

Peizhi Xu

Jiacheng Guo

Jinshan Liu

Xuejun Xia

An-Ping Zeng

Xingguo Xing

Renyun Wang

Jian Bo

Qie Yin

Dongliang Chen

Sujuan Sun

Dongxu Ma

Minsheng Wu

Zeming Zhang

Xuezhong Qiu

Baichuan Wang

Yueran Dang

Feng Geng

Zhufang Shen

Yueteng Chen

Mingzhi Xie

Jianfa Wu

Shou Ma

Yinghao Du

Zengyao Pang

Yubao Gao

Wulin Tian

Tianfei Wang

Buhe Han

Qian Qi

Xiaochun Liu

Sichen Xu

Xiaoyue Liu

Mi Cao

Cunchao Wei

A system determines a timing matrix corresponding to inference times taken for a number of machine learning (ML) models to be executed by a number of processing resources of a computing device. The processing resources includes at least a first and a second type of processing resources. The system applies a service-specific model-first scheduling scheme or a service-specific hardware-first scheduling scheme to obtain corresponding service-specific mappings. The system determines a best mapping from the corresponding service-specific mappings. The system schedules each of the ML models to a corresponding processing resource from the processing resources according to the best mapping. The system executes the ML models using corresponding mapped processing resources.

Scheduling ML services and models with heterogeneous resources

A system obtains a performance profile corresponding to times taken to perform an inferencing by a machine learning (ML) model using a different number of processing resources from a plurality of processing resources. The system determines one or more groupings of processing resources from the plurality of processing resources, each grouping includes one or more partitions. The system calculates performance speeds corresponding to each grouping based on the performance profile. The system determines a grouping having a best performance speed from the calculated performance speeds. The system partitions the processing resources based on the determined grouping to perform the inferencing.

System on chip parallel computing of ML services and applications for partitioning resources based on the inference times

An autonomous mobile device and a control method are disclosed. The autonomous mobile device includes a main body and an imaging device, a detecting device, a light source assembly, and a controller disposed at the main body. The imaging device is configured to acquire image information in a predetermined direction of the main body. The predetermined direction includes an upward direction of the main body. The detecting device is configured to detect obstacle information in the upward direction of the main body. A light outputting direction of the light source assembly includes the upward direction of the main body. The controller is configured to, when the detecting device detects the obstacle within a predetermined distance range in the upward direction of the main body, control the light source assembly to turn on to provide illumination for the imaging device.

Autonomous mobile device and control method thereof

A cleaning robot is provided. The cleaning robot includes a motion device configured to move the cleaning robot in an environment, and an exterior housing. The cleaning robot also includes a motion sensor configured to obtain parameters relating to a motion of the cleaning robot, and a storage device configured to store data including at least one of the one or more images or the one or more motion parameters. The cleaning robot also includes a camera configured to capture one or more images of the environment. The camera is disposed internal to a slant surface located at a front end of the exterior housing and is pivotable relative to the slant surface. The slant surface inclines upwardly with respect to a forward moving direction. The camera and the slant surface face substantially a same direction. A direction of the camera and the forward moving direction form an acute angle.

Cleaning robot

A method for processing a map of a closed space is provided. The method includes obtaining a plurality of self-locations of a mobile device. The method also includes obtaining a plurality of obstacle locations. The method also includes obtaining a boundary profile based on the obstacle locations, the boundary profile being a closed line formed by sequentially connecting multiple adjacent obstacle locations. The method also includes obtaining a self-circle based on the self-locations of the mobile device, the self-circle being a closed line formed by sequentially connecting multiple adjacent self-location of the mobile device. The method also includes determining the boundary profile as an outer profile when the boundary profile encircles the self-circle at an outer side of the self-circle. The method further includes processing the map of the closed space based on the outer profile.

Method for processing map of closed space, apparatus, and mobile device

A method includes obtaining a map of a closed space. The method also includes obtaining trajectory points. The method also includes processing the trajectory points, and recognizing a correct door in the closed space based on the map and a result of processing the trajectory points. The method also includes dividing the closed space into zones based on the correct door and the map.

Method, apparatus for zone division of closed space and mobile device

A docking method executable by a mobile device is provided. The docking method includes obtaining a stored target location of a docking station, and navigating to the target location. The docking method also includes: during the navigation and/or at the target location, based on a determination that a guidance signal is not detected, performing a regional search. The docking method also includes: during the navigation, or at the target location, or during the regional search, based on a determination that the guidance signal is detected, moving, under the guidance of the guidance signal, to the docking station. Performing the regional search includes determining a basic search zone, searching for the guidance signal while moving along boundaries of the basic search zone, and based on a determination that the guidance signal has not been detected when a termination condition is satisfied, terminating the regional search.

Mobile device docking method and mobile device

Provided are a method for processing data, an electronic device and a storage medium, which relate to the field of deep learning and data processing. The method may include: multiple target operators of a target model are acquired; the multiple target operators are divided into at least one operator group, according to an operation sequence of each of the multiple target operators in the target model, wherein at least one target operator in each of the at least one operator group is operated by the same processor and is operated within the same target operation period; and the at least one operator group is output.

Method for processing data, electronic device and storage medium

A method executable by an autonomous mobile device includes moving in a work environment, obtaining environmental data acquired by a sensing device, and determining whether the sensing device is in a suspected ineffective state based on the environmental data. The method also includes based on a determination that the sensing device is in the suspected ineffective state, rotating at a same location for a first predetermined spin angle. The method also includes obtaining an estimated rotation angle based on one or more motion parameters acquired by a dead reckoning sensor, comparing the estimated rotation angle with the first predetermined spin angle, and based on a determination that a difference between the estimated rotation angle and the first predetermined spin angle is greater than a first predetermined threshold value, executing escape instructions to move backwardly for a first predetermined distance and move along a curve or a folded line.

Method and apparatus for autonomous mobile device

This application discloses a page displaying method and apparatus, an electronic device and a storage medium, relating to the technical field of computer vision, including: extracting a theme color from an image of a page to be displayed, and determining a coloration attribute of the theme color, where the coloration attribute is a chromatic attribute, or the coloration attribute is a non-chromatic attribute; and adjusting the page to be displayed according to the coloration attribute and displaying the page. Various pages are adjusted and displayed based on different coloration attributes of the theme color, thus avoiding a disadvantage related to lack of flexibility caused by filtering the theme color by a preset filtering range in related art, realizing diversity and flexibility in displaying for various pages, avoiding the disadvantage of distortion of images caused by simple filtering, and improving the reliability and accuracy of page display.

Page displaying method and apparatus, electronic device and storage medium

The present application discloses a speech recognition method, apparatus, device and readable storage medium, and relates to the technical field of artificial intelligence. A specific implementation includes: an electronic device recognizes a speech signal to obtain a first text; if a first pinyin sequence corresponding to the first text exists in the database, the electronic device uses a correct text corresponding to the first pinyin sequence as the speech recognition result; otherwise, the electronic device performs a fuzzy matching on the first pinyin sequence to obtain multiple second pinyin sequences and second texts corresponding to the second pinyin sequences, and selects the speech recognition result from the multiple second texts.

Speech recognition method, apparatus, device and readable storage medium

A method includes acquiring an image through a visual sensor during a movement of a mobile device. The method includes matching the image with key frames stored in a key frame database. The key frames are created based on two-dimensional coordinates of feature points included in a plurality of images previously acquired through the visual sensor. The method also includes computing a visual relative pose based on two-dimensional coordinates of matching feature points included in both of the image and the one or more key frames that have been matched with the image. The method also includes computing relevant information of the visual relative pose based on the two-dimensional coordinates of the matching feature points. The method further includes updating an absolute pose of the mobile device and a map based on the relevant information of the visual relative pose and relevant information of a dead reckoning based relative pose.

Methods, apparatus, and systems for localization and mapping

Provided is a method for producing 1,3-propanediol by means of fermentation of a recombinant microorganism. First, a recombinant microorganism is provided; the recombinant microorganism can overexpress acetyl-CoA carboxylase genes: accBC and accDA, a malonyl-CoA synthetase gene: mcr, a 3-hydroxypropionyl-CoA synthetase gene: pcs, a 3-hydroxypropionyl-CoA reductase gene: pduP, and a 1,3-propanediol reductase gene: yqhD. The recombinant microorganism is subjected to fermentation culture in a flask or fermentor using glucose ad as raw material to obtain the 1,3-propanediol. The recombinant microorganism can utilize low-cost glucose, sucrose, molasses, xylose and the like as raw material in the fermentation process, without additional expensive vitamin B12. Thus, cost of the production is significantly reduced, and there is a promising prospect in market.

Method for producing 1,3-propanediol by fermentation of a recombinant microorganism

Growing community of inventors

Beijing, China

Zhen Chen