公开(公告)号：WO2020047823A1

公开(公告)日：2020-03-12

申请号：PCT/CN2018/104539

申请日：2018-09-07

Applicant: INTEL CORPORATION ,  ZHANG, Yu ,  LE, Huifeng ,  CHUANG, Richard ,  WERNER, Metz, Jr. ,  HAN, Heng Juen ,  ZHANG, Ning ,  SHAO, Wenjian ,  HE, Ke

Inventor： ZHANG, Yu ,  LE, Huifeng ,  CHUANG, Richard ,  WERNER, Metz, Jr. ,  HAN, Heng Juen ,  ZHANG, Ning ,  SHAO, Wenjian ,  HE, Ke

IPC: G06N3/063

Abstract: Processes and systems are disclosed. The processes and systems are arranged to apply convolution for a CNN where the CNN is simplified using sparse techniques, quantization techniques or both sparse and quantization techniques. A location vector (LV) table is provided to record the coordinates of non-zero weights. A look up table is provided to recover the real weight value from the weight identification. Convolution is applied by retrieving the coordinates of the next non-zero weight and the associated real weight value and by accumulating the multiplication of the real weight value and the input value across the input activation plane.

公开(公告)号：WO2022115991A1

公开(公告)日：2022-06-09

申请号：PCT/CN2020/133084

申请日：2020-12-01

Applicant: INTEL CORPORATION ,  YAO, Anbang ,  KANG, Yangyuxuan ,  WANG, Shandong ,  LU, Ming ,  CHEN, Yurong ,  SHAO, Wenjian ,  WANG, Yikai ,  XU, Haojun ,  YU, Chao ,  WONG, Chong

Inventor： YAO, Anbang ,  KANG, Yangyuxuan ,  WANG, Shandong ,  LU, Ming ,  CHEN, Yurong ,  SHAO, Wenjian ,  WANG, Yikai ,  XU, Haojun ,  YU, Chao ,  WONG, Chong

IPC: G06T7/207

Abstract: Techniques related to 3D pose estimation from a 2D input image are discussed. Such techniques include incrementally adjusting an initial 3D pose generated by applying a lifting network to a detected 2D pose in the 2D input image by projecting each current 3D pose estimate to a 2D pose projection, applying a residual regressor to features based on the 2D pose projection and the detected 2D pose, and combining a 3D pose increment from the residual regressor to the current 3D pose estimate.

公开(公告)号：WO2017166119A1

公开(公告)日：2017-10-05

申请号：PCT/CN2016/077856

申请日：2016-03-30

Applicant: INTEL CORPORATION ,  LE, Huifeng ,  SHAO, Wenjian ,  ZHANG, Yu ,  YANG, Shao-wen ,  HAN, Heng Juen ,  ZHANG, Xiaowen

Inventor： LE, Huifeng ,  SHAO, Wenjian ,  ZHANG, Yu ,  YANG, Shao-wen ,  HAN, Heng Juen ,  ZHANG, Xiaowen

IPC: H04N7/18 ,  H04L29/08

CPC classification number: H04L47/70 ,  H04L29/08 ,  H04L43/16 ,  H04L67/025 ,  H04L67/10 ,  H04L67/1002 ,  H04L67/22 ,  H04N7/181

Abstract: Techniques are provided for adaptive distribution of video analysis workload over a network of video processor nodes. The nodes may include, for example, internet protocol (IP) cameras, video recorders and/or data centers. The network may also include a management system configured to assign video analysis tasks to the nodes based on the node resources and predictive modelling of the node workload. The management system may re-distribute the tasks based on performance monitoring. Some assigned tasks may be bound to the node while other tasks may be transferrable, by the node, to other nodes. The nodes may be configured to determine which of the transferrable tasks will be locally executed or transferred based on a check of resource usage against a usage policy that specifies thresholds for the determinations. The nodes may be configured to transmit video analysis packets, including image data, analysis completion status and analysis results, to other nodes.

Abstract translation: 提供了用于在视频处理器节点的网络上自适应分配视频分析工作负载的技术。 节点可以包括例如网际协议（IP）摄像机，录像机和/或数据中心。 网络还可以包括管理系统，该管理系统被配置为基于节点资源和节点工作负载的预测建模将视频分析任务分配给节点。 管理系统可以根据性能监控重新分配任务。 一些分配的任务可以绑定到节点，而其他任务可以由节点转移到其他节点。 节点可以被配置成基于对指定用于确定的阈值的使用策略的资源使用的检查来确定哪些可传输任务将被本地执行或传输。 节点可以被配置为将视频分析包（包括图像数据，分析完成状态和分析结果）传送到其他节点。