Patent US 11,620,766

Patent №

US 11,620,766

Granted

—

Owner

—

Lab

—

AI components

ml · vision · hardware

Assignment

None on record

Dataset

AIPD

2023_r1 edition

Application

17344639

In an example, an apparatus comprises logic, at least partially including hardware logic, to implement a lossy compression algorithm which utilizes a data transform and quantization process to compress data in a convolutional neural network (CNN) layer.

Machine learning Vision AI hardwareG06T 9/002G06N 3/044G06N 3/0442G06N 3/045G06N 3/0464G06N 3/047G06N 3/048G06N 3/0495+6 more

Google Patents

AI classification

Machine learning1.00

Vision1.00

AI hardware1.00

Planning0.19

Evolutionary computation0.01

Speech0.00

Knowledge representation0.00

Natural language0.00

Cites (53)

NEURAL NETWORK REDUCTION DEVICE, NEURAL NETWORK REDUCTION METHOD, AND STORAGE MEDIUMUS 11,449,734 · 2022 SYSTEMS AND METHODS FOR TRANSLATING WITH LIMITED ATTENTIONUS 11,410,015 · 2022 MACHINE-LEARNED DESKING VEHICLE RECOMMENDATIONUS 11,386,161 · 2022 SYSTEMS AND METHODS FOR CROSS-PRODUCT MALWARE CATEGORIZATIONUS 11,023,580 · 2021 DISTRIBUTABLE EVENT PREDICTION AND MACHINE LEARNING RECOGNITION SYSTEMUS 11,010,691 · 2021 SPARSE CONVOLUTIONAL NEURAL NETWORK ACCELERATORUS 10,891,538 · 2021 SPARSE CONVOLUTIONAL NEURAL NETWORK ACCELERATORUS 10,860,922 · 2020 SYSTEM AND METHOD FOR ANOMALY DETECTION IN DYNAMICALLY EVOLVING DATA USING RANDOM NEURAL NETWORK DECOMPOSITIONUS 10,692,004 · 2020 DETECTION METHOD AND INFORMATION PROCESSING DEVICEUS 10,572,512 · 2020 SPARSE CONVOLUTIONAL NEURAL NETWORK ACCELERATORUS 10,528,864 · 2020 COMPRESSED RECURRENT NEURAL NETWORK MODELSUS 10,515,307 · 2019 PREDICTABLE ORGANIC TILE LAYOUTUS 10,452,749 · 2019 PROCESSOR WITH MEMORY ARRAY OPERABLE AS EITHER LAST LEVEL CACHE SLICE OR NEURAL NETWORK UNIT MEMORYUS 10,430,706 · 2019 DECOMPOSING CONVOLUTION OPERATION IN NEURAL NETWORKSUS 10,402,720 · 2019 REAL-TIME CO-CHANNEL INTERFERENCE SUPPRESSIONUS 10,334,455 · 2019 AUGMENTED GENERALIZED DEEP LEARNING WITH SPECIAL VOCABULARYUS 10,210,860 · 2019 RECOGNITION, REIDENTIFICATION AND SECURITY ENHANCEMENTS USING AUTONOMOUS MACHINESUS 10,108,850 · 2018 Phoneme-Expert Assisted Speech Recognition & Re-synthesisUS 9,792,897 · 2017 VECTOR-BASED SHOCK INDICATIONUS 9,782,093 · 2017 Rank-Constrained Neural NetworksUS 9,767,410 · 2017 ENHANCED POWER METHOD ON AN ELECTRONIC DEVICEUS 9,684,538 · 2017 ELECTRICAL LOAD IDENTIFICATION DURING AN ON EVENT USING A SUPPORT VECTOR MACHINE CLASSIFIERUS 9,606,154 · 2017 CONTENT ADAPTIVE TELECINE AND INTERLACE REVERSERUS 9,521,358 · 2016 Neural Signal Processing and/or Interface Methods, Architectures, Apparatuses, and DevicesUS 9,477,640 · 2016 Compression Using an Iterative Data Derived Basis SetUS 9,479,799 · 2016

and 28 more

Related research

A Dynamic Codec with Adaptive Quantization for Convolution Neural NetworkOpenAlex A Hybrid CNN Compression Approach for Hardware AccelerationOpenAlex Quantization-Based Optimization Algorithm for Hardware Implementation of Convolution Neural NetworksOpenAlex Memory-Efficient Compression Based on Least-Squares Fitting in Convolutional Neural Network AcceleratorsOpenAlex Logic-Based Neural Network for Image Compression ApplicationsSemantic Scholar

Full text

Google Patents