公开(公告)号：US09996764B2

公开(公告)日：2018-06-12

申请号：US15306983

申请日：2014-04-29

Applicant: Institute of Automation Chinese Academy of Sciences

Inventor： Jian Cheng ,  Cong Leng ,  Jiaxiang Wu ,  Hanqing Lu

IPC: G06K9/00 ,  G06K9/46 ,  G06F17/30 ,  G06K9/62

CPC classification number: G06K9/4633 ,  G06F17/30256 ,  G06F17/30268 ,  G06F17/30271 ,  G06K9/4671 ,  G06K9/6215

Abstract: An image matching method based on cascaded binary encoding includes using a hashing look-up with multiple hashing tables to coarsely filter candidate key-points in an image to produce a candidate subset of key-points, projecting the candidate subset into a high-dimensional Hamming space, and building a “Hamming distance-memory address” hashing table. An optimal matching key-point is discovered by querying the hashing table. The image matching method has high processing speed and matching quality, which can be used for efficient and accurate image matching.

公开(公告)号：US20180247180A1

公开(公告)日：2018-08-30

申请号：US15753520

申请日：2015-08-21

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jian Cheng ,  Jiaxiang Wu ,  Cong Leng ,  Hanqing Lu

IPC: G06N3/04 ,  G06N3/08 ,  G06F17/16

CPC classification number: G06N3/04 ,  G06F17/16 ,  G06K9/4628 ,  G06K9/6243 ,  G06K9/6274 ,  G06N3/0454 ,  G06N3/08 ,  G06N20/00

Abstract: An acceleration and compression method for a deep convolutional neural network based on quantization of a parameter provided by the present application comprises: quantizing the parameter of the deep convolutional neural network to obtain a plurality of subcode books and respective corresponding index values of the plurality of subcode books; acquiring an output feature map of the deep convolutional neural network according to the plurality of subcode books and respective corresponding index values of the plurality of subcode books. The present application may implement the acceleration and compression for a deep convolutional neural network.

公开(公告)号：US10970617B2

公开(公告)日：2021-04-06

申请号：US15753520

申请日：2015-08-21

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jian Cheng ,  Jiaxiang Wu ,  Cong Leng ,  Hanqing Lu

IPC: G06N3/04 ,  G06K9/62 ,  G06K9/46 ,  G06N20/00 ,  G06F17/16 ,  G06N3/08

Abstract: An acceleration and compression method for a deep convolutional neural network based on quantization of a parameter provided by the present application comprises: quantizing the parameter of the deep convolutional neural network to obtain a plurality of subcode books and respective corresponding index values of the plurality of subcode books; acquiring an output feature map of the deep convolutional neural network according to the plurality of subcode books and respective corresponding index values of the plurality of subcode books. The present application may implement the acceleration and compression for a deep convolutional neural network.