公开(公告)号：US20230006611A1

公开(公告)日：2023-01-05

申请号：US17857132

申请日：2022-07-04

Applicant: MediaTek Inc.

Inventor： Po-Yu Chen ,  Hao Chen ,  Yi-Min Tsai ,  Hao Yun Chen ,  Hsien-Kai Kuo ,  Hantao Huang ,  Hsin-Hung Chen ,  Yu Hsien Chang ,  Yu-Ming Lai ,  Lin Sen Wang ,  Chi-Tsan Chen ,  Sheng-Hong Yan

IPC: H03F1/32 ,  H03F1/02 ,  G06K9/62 ,  G06N3/063

Abstract: A compensator compensates for the distortions of a power amplifier circuit. A power amplifier neural network (PAN) is trained to model the power amplifier circuit using pre-determined input and output signal pairs that characterize the power amplifier circuit. Then a compensator is trained to pre-distort a signal received by the PAN. The compensator uses a neural network trained to optimize a loss between a compensator input and a PAN output, and the loss is calculated according to a multi-objective loss function that includes one or more time-domain loss function and one or more frequency-domain loss functions. The trained compensator performs signal compensation to thereby output a pre-distorted signal to the power amplifier circuit.

公开(公告)号：US20230064692A1

公开(公告)日：2023-03-02

申请号：US17846007

申请日：2022-06-22

Applicant: MediaTek Inc.

Inventor： Hao Yun Chen ,  Min-Hung Chen ,  Min-Fong Horng ,  Yu-Syuan Xu ,  Hsien-Kai Kuo ,  Yi-Min Tsai

IPC: G06N3/04 ,  G06N3/08 ,  G06N3/063 ,  G06N7/00

Abstract: According to a network space search method, an expanded search space is partitioned into multiple network spaces. Each network space includes a plurality of network architectures and is characterized by a first range of network depths and a second range of network widths. The performance of the network spaces is evaluated by sampling respective network architectures with respect to a multi-objective loss function. The evaluated performance is indicated as a probability associated with each network space. The method then identifies a subset of the network spaces that has the highest probabilities, and selects a target network space from the subset based on model complexity.

公开(公告)号：US11742875B1

公开(公告)日：2023-08-29

申请号：US17724849

申请日：2022-04-20

Applicant: MediaTek Inc.

Inventor： Hsien-Kai Kuo ,  Huai-Ting Li ,  Shou-Yao Tseng ,  Po-Yu Chen

IPC: H03M7/00 ,  H03M7/30 ,  G06F9/50 ,  G06N3/04 ,  G06N3/08 ,  G06N3/063

CPC classification number: H03M7/30 ,  G06F9/5027 ,  G06N3/04 ,  G06N3/063 ,  G06N3/08

Abstract: Floating-point numbers are compressed for neural network computations. A compressor receives multiple operands, each operand having a floating-point representation of a sign bit, an exponent, and a fraction. The compressor re-orders the operands into a first sequence of consecutive sign bits, a second sequence of consecutive exponents, and a third sequence of consecutive fractions. The compressor then compresses the first sequence, the second sequence, and the third sequence to remove at least duplicate exponents. As a result, the compressor can losslessly generate a compressed data sequence.

公开(公告)号：US20230196526A1

公开(公告)日：2023-06-22

申请号：US17552912

申请日：2021-12-16

Applicant: MediaTek Inc.

Inventor： Yu-Syuan Xu ,  Yu Tseng ,  Shou-Yao Tseng ,  Hsien-Kai Kuo ,  Yi-Min Tsai

IPC: G06T5/50 ,  G06T5/00 ,  G06N3/04

CPC classification number: G06T5/50 ,  G06T5/001 ,  G06N3/04

Abstract: A system stores parameters of a feature extraction network and a refinement network. The system receives an input including a degraded image concatenated with a degradation estimation of the degraded image; performs operations of the feature extraction network to apply pre-trained weights to the input to generate feature maps; and performs operations of the refinement network including a sequence of dynamic blocks. One or more of the dynamic blocks dynamically generates per-grid kernels to be applied to corresponding grids of an intermediate image output from a prior dynamic block in the sequence. Each per-grid kernel is generated based on the intermediate image and the feature maps.