-
公开(公告)号：US20220222523A1
公开(公告)日：2022-07-14
申请号：US17206164
申请日：2021-03-19
Inventor： Hoi Jun YOO , Dong Hyeon HAN
Abstract: Disclosed herein are an apparatus and method for training a low-bit-precision deep neural network. The apparatus includes an input unit configured to receive training data to train the deep neural network, and a training unit configured to train the deep neural network using training data, wherein the training unit includes a training module configured to perform training using first precision, a representation form determination module configured to determine a representation form for internal data generated during an operation procedure for the training and determine a position of a decimal point of the internal data so that a permissible overflow bit in a dynamic fixed-point system varies randomly, and a layer-wise precision determination module configured to determine precision of each layer during an operation in each of a feed-forward stage and an error propagation stage and automatically change the precision of a corresponding layer based on the result of determination.
-
公开(公告)号：US20210056427A1
公开(公告)日：2021-02-25
申请号：US16988737
申请日：2020-08-10
Inventor： Hoi Jun YOO , Dong Hyeon HAN
Abstract: Disclosed herein are an apparatus and method for training a deep neural network. An apparatus for training a deep neural network including N layers, each having multiple neurons, includes an error propagation processing unit configured to, when an error occurs in an N-th layer in response to initiation of training of the deep neural network, determine an error propagation value for an arbitrary layer based on the error occurring in the N-th layer and directly propagate the error propagation value to the arbitrary layer, a weight gradient update processing unit configured to update a forward weight for the arbitrary layer based on a feed-forward value input to the arbitrary layer and the error propagation value in response to the error propagation value, and a feed-forward processing unit configured to, when update of the forward weight is completed, perform a feed-forward operation in the arbitrary layer using the forward weight.
-
公开(公告)号：US20240362848A1
公开(公告)日：2024-10-31
申请号：US18628865
申请日：2024-04-08
Inventor： Hoi Jun YOO , Dong hyeon HAN
Abstract: Provided is a 3D rendering accelerator based on a DNN trained using a weight of the DNN using a plurality of 2D photos obtained by imaging the same object from several directions and then configured to perform 3D rendering using the same, the 3D rendering accelerator including a VPC configured to create an image plane for a 3D rendering target from a position and a direction of an observer, divide the image plane into a plurality of tile units, and then perform brain imitation visual recognition on the divided tile-unit images to determine to reduce a DNN inference range, an HNE including a plurality of NEs having different operational efficiencies and configured to accelerate DNN inference by dividing and allocating tasks, and a DNNA core configured to generate selection information for allocating each task to one of the plurality of NEs based on a sparsity ratio.
-
4.发明公开公开(公告)号：US20240078282A1
公开(公告)日：2024-03-07
申请号：US18139952
申请日：2023-04-27
Inventor： Hoi Jun YOO , Dong Seok IM
CPC classification number: G06F17/16 , G06T7/521 , H04N19/174
Abstract: Disclosed is a conjugate gradient acceleration apparatus using band matrix compression in depth fusion technology including a band matrix conversion unit configured to convert an adjacency matrix for correcting depth data acquired from data of an image sensor through deep learning based on depth information acquired from a depth sensor into a band matrix using rows of the adjacency matrix as addresses of query points and columns of the adjacency matrix as the nearest neighbors at the query points, a band matrix compression unit configured to mark an index on each band in order to compress the band matrix and to compress data, a memory unit configured to store tile data of the band matrix, and a band matrix calculation unit configured to perform computation of the band matrix and a transposed band matrix or computation of a symmetric band matrix with respect to the band matrix and a vector.
-
5.发明申请公开(公告)号：US20230098672A1
公开(公告)日：2023-03-30
申请号：US17574501
申请日：2022-01-12
Inventor： Hoi Jun YOO , So Yeon KIM
IPC: G06N3/08 , G06V10/82 , G06V10/776
Abstract: Disclosed is an energy-efficient retraining method of a generative neural network for domain-specific optimization, including (a) retraining, by a mobile device, a pretrained generative neural network model with respect to some data of a new user dataset, (b) comparing, by the mobile device, the pretrained generative neural network model and a generative neural network model retrained for each layer with each other in terms of a relative change rate of weights, (c) selecting, by the mobile device, specific layers having high relative change rate of weights, among layers of the pretrained generative neural network model, as layers to be retrained, and (d) performing, by the mobile device, weight update for only the layers selected in step (c), wherein only some of all layers are selected and trained in a retraining process that requires a large amount of operation, whereby rapid retraining is performed in the mobile device.
-
Patent Agency Ranking
公开(公告)号：US20230376756A1
公开(公告)日：2023-11-23
申请号：US18199995
申请日：2023-05-22
Inventor： Hoi Jun YOO , Dong Seok IM
IPC: G06N3/08
CPC classification number: G06N3/08
Abstract: Disclosed is a 3D point cloud-based deep learning neural network acceleration apparatus including a depth image input unit configured to receive a depth image, a depth data storage unit configured to store depth data derived from the depth image, a sampling unit configured to sample the depth image in units of a sampling window having a predetermined first size, a grouping unit configured to generate a grouping window having a predetermined second size and to group inner 3D point data by grouping window, and a convolution computation unit configured to separate point feature data and group feature data, among channel-direction data of 3D point data constituting the depth image, to perform convolution computation with respect to the point feature data and the group feature data, to sum the results of convolution computation by group grouped by the grouping unit, and to derive the final result.
-
公开(公告)号：US20230195420A1
公开(公告)日：2023-06-22
申请号：US17741509
申请日：2022-05-11
Inventor： Hoi Jun YOO , Ju Hyoung LEE
CPC classification number: G06F7/5443 , G06F7/4876 , G06F7/485 , G06F7/49915 , G06F5/012 , G06N3/04
Abstract: Disclosed herein are a floating-point computation apparatus and method using Computing-in-Memory (CIM). The floating-point computation apparatus performs a multiply-and-accumulation operation on pieces of input neuron data represented in a floating-point format, and includes a data preprocessing unit configured to separate and extract an exponent and a mantissa from each of the pieces of input neuron data, an exponent processing unit configured to perform CIM on input neuron exponents, which are exponents separated and extracted from the input neuron data, and a mantissa processing unit configured to perform a high-speed computation on input neuron mantissas, separated and extracted from the input neuron data, wherein the exponent processing unit determines a mantissa shift size for a mantissa computation and transfers the mantissa shift size to the mantissa processing unit, and the mantissa processing unit normalizes a result of the mantissa computation and transfers a normalization value to the exponent processing unit.
-
公开(公告)号：US20180004287A1
公开(公告)日：2018-01-04
申请号：US15542933
申请日：2015-12-04
Inventor： Hoi Jun YOO , In Joon Hong , Kyeong Ryeol BONG , Jun Young PARK
IPC: G06F3/01 , G06T7/246 , G02B27/01 , G06F3/0481
CPC classification number: G06F3/013 , G02B27/0093 , G02B27/017 , G02B2027/014 , G02B2027/0141 , G02B2027/0178 , G02B2027/0187 , G06F1/163 , G06F3/011 , G06F3/017 , G06F3/0481 , G06F3/0482 , G06F3/04842 , G06T7/248
Abstract: A method for providing a user interface through a head mounted display using eye recognition and bio-signals comprises the steps of: (a) moving a cursor to a particular location at which a user gazes by referencing the eye information obtained from a first eyeball that is one of the eyeballs of the user through a camera module when the user gazes at a particular location on an output screen; and (b) supporting in order to provide detailed selection items corresponding to an entity when a certain entity exists in the certain position by referencing the movement information obtained from the eyelid corresponding to a second eyeball that is one of the eyeballs of the user through a bio-signal acquisition module.
-
9.发明申请公开(公告)号：US20250103241A1
公开(公告)日：2025-03-27
申请号：US18644111
申请日：2024-04-24
Inventor： Hoi Jun YOO , Sang Jin KIM
IPC: G06F3/06
Abstract: A DRAM is configured using a triple-mode memory cell that supports a computation mode, a memory mode, and a data conversion mode by one cell and converts modes as necessary, and an AI accelerator using the same is provided, so that a dataflow may be reconfigured according to a structure and a size of an AI neural network (so-called deep neural network) to be trained.
-
公开(公告)号：US20240330664A1
公开(公告)日：2024-10-03
申请号：US18381218
申请日：2023-10-18
Inventor： Hoi Jun YOO , Dong Seok IM
IPC: G06N3/063
CPC classification number: G06N3/063
Abstract: A signed bit slice generator includes a divider configured to divide input data, which is 2's complement data having N (where N is a natural number)-bit precision, and divide remaining bits excluding a sign bit of the input data into a predetermined number of bit slices, a sign bit adder configured to add a sign bit to each of the bit slices, a sign value setter configured to set a sign bit of an MSB slice among the bit slices to a sign value of the input data and to set sign bits of the remaining bit slices to positive sign values, and a sparse data compressor configured to perform sparse data compression on each of the signed bit slices, thereby generating a predetermined number of signed bit slices having the same number of bits where each bit slice includes a sign bit.
-
-
-
-
-
-
-
-
-
Search Results
14
records
(took 0.024 seconds)
