IP Library Granted Patent US 12,725,032
Granted Patent B2
US 12,725,032 · App. 17/473,685 · Granted Sep 1, 2026

Systems and methods for distributed data analytics

Inventors: Lars Oleson (Bondi, AU); Shivanthan Yohanandan (Pakenham, AU); Ryan Mccrea (Etobicoke, CA); Deepa Lakshmi Chandrasekharan (Strathfiled, IN); Sabina Pokhrel (Kathmandu, NP); Yousef Rabi (Peakhurst, AU); Zhenhua Zhang (Chatswood, AU); Priyadharshini Devanand (Lidcombe, IN); Bernardo Rodeiro Croll (Kingsford, AU); James J. Meyer (Van Meter, IA)
Assignee: Xailient
G06N3/08G06F18/2148G06F18/217G06F18/28G06F18/40G06V10/774G06V10/82G06V10/945G06V40/16G06T2207/20084G06V10/462G06V40/161
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,725,032
App. No.
17/473,685
Granted
Sep 1, 2026
Kind
B2
Abstract

The invention provides systems and method for generating device-specific artificial neural network (ANN) models for distribution across user devices. Sample datasets are collected from devices in a particular environment or use case and include predictions by device-specific ANN models executing the user devices. The received datasets are used with existing datasets and stored ANN models to generate updated device-specific ANN models from each of the stored instances of the device ANN models based on the training data.

Claims (27)

1 . A method for optimizing the execution of device-specific trained artificial neural network (ANN) models on devices, the method comprising:

receiving, by a processor, a first trained ANN model and a second trained ANN model, wherein the first trained ANN model and the second ANN model each execute different inferences on input data and are trained using the same training dataset comprising the same input samples;

merging the first trained ANN model, the second trained ANN model, and control flow execution instructions into a combined software package; and

deploying the combined software package to a hardware component of an edge device through a network; and

executing, by the hardware component of the edge device, the combined software package, wherein deploying the combined software package comprises selecting either or both of the first and second trained ANN models for performing object detection based on the control flow instructions according to analytics criteria.

2 . The method of claim 1 , wherein output of the first ANN model serves as input to the second ANN model.

3 . The method of claim 1 , wherein the first trained ANN model and second trained ANN model each comprise use case data.

4 . The method of claim 1 , further comprising generating a parent ANN as a meta-architecture based on the first ANN model architecture and the second ANN model architecture, and the meta-architecture is delivered to the edge device such that it executes as a single ANN model.

5 . The method of claim 1 , wherein the edge device comprises a camera.

6 . The method of claim 5 , wherein execution of the first ANN model and second ANN model on the camera identifies an object of interest in an image file captured on the camera.

7 . A system for optimizing the execution of device-specific trained artificial neural network (ANN) models on edge devices, the system comprising:

one or more processors; and

a memory coupled with the one or more processors wherein the one or more processors executes computer-executable instructions stored in the memory, that when executed:

identify a first trained ANN model and a second trained ANN model, wherein the first trained ANN model and the second trained ANN model each execute different inferences on input data and are trained using the same training dataset comprising the same input samples;

merge the first trained ANN model, the second trained ANN model, and control flow execution instructions into a combined software package; and

deploy, by a distribution module, the combined software package to a hardware component of an edge device through a network; and

execute, by the hardware component of the edge device, the combined software package, wherein the edge device is configured to select either or both of the first and second trained ANN models for performing object detection based on the control flow instructions included in the wrapper function according to analytics criteria.

8 . The system of claim 7 , wherein the output of the first ANN model serves as input to the second ANN model.

9 . The system of claim 7 , wherein the first trained ANN model and second trained ANN model each comprise respective use case data.

10 . The system of claim 7 , wherein execution of the instructions further generates a parent ANN as a meta-architecture based on the first ANN model architecture and the second ANN model architecture, and the meta-architecture is delivered to the edge device such that it executes as a single ANN model.

11 . The system of claim 7 , wherein the edge device comprises a camera.

12 . The system of claim 11 , wherein execution of the first ANN model and second ANN model on the camera identifies an object of interest in an image file captured on the camera.

13 . A non-transitory computer-readable storage medium storing computer-readable instructions that, when executed by a computer, cause the computer to perform a method, the method comprising:

receiving, by a processor, a first trained ANN model and a second trained ANN model, wherein the first trained ANN model and the second trained ANN model each execute different inferences on input data and are trained using the same training dataset comprising the same input samples;

merging the first trained ANN model, the second trained ANN model, and control flow execution instructions into a combined software package;

communicating the combined software package to a hardware component of an edge device through a network; and

deploying, by the hardware component of the edge device, the combined software package, wherein the edge device is configured to select either or both of the first and second trained ANN models for performing object detection based on the control flow instructions included in the wrapper function according to analytics criteria.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 17, 2021
From: MCCREA, RYAN; CHANDRASEKHARAN, DEEPA LAKSHMI; POKHREL, SABINA; RABI, YOUSEF; ZHANG, ZHENHUA; DEVANAND, PRIYADHARSHINI; RODEIRO CROLL, BERNARDO; OLESON, LARS; YOHANANDAN, SHIVANTHAN
To: XAILIENT
Reel/Frame 057515/0757 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 17, 2021
From: MEYER, JAMES J.
To: XAILIENT
Reel/Frame 057538/0385 →
Continuity (3)
Continuation 17314465 · May 7, 2021
Provisional Application 63021735 · May 8, 2020
Related Publication 20210406607A1 · Dec 30, 2021
References Cited (72)
US 7599894B2 · Owechko et al. · 2009 [cited by applicant]
US 9354709B1 · Heller et al. · 2016 [cited by applicant]
US 10007867B2 · Yu et al. · 2018 [cited by applicant]
US 10979625B2 · Lee et al. · 2021 [cited by applicant]
US 11423616B1 · Simon Kreuz et al. · 2022 [cited by applicant]
US 11562542B2 · Zhang · 2023 [cited by examiner]
US 11669332B1 · Aiyer et al. · 2023 [cited by applicant]
US 11701090B2 · Moctezuma de la Barrera et al. · 2023 [cited by applicant]
US 20050147297A1 · McLaughlin · 2005 [cited by examiner]
US 20130176430A1 · Zhu et al. · 2013 [cited by applicant]
US 20130294651A1 · Zhou et al. · 2013 [cited by applicant]
US 20140355821A1 · Solem et al. · 2014 [cited by applicant]
US 20150089399A1 · Megill et al. · 2015 [cited by applicant]
US 20170286809A1 · Pankanti · 2017 [cited by examiner]
US 20180060240A1 · Kadav · 2018 [cited by examiner]
US 20180204088A1 · Chen et al. · 2018 [cited by applicant]
US 20180285682A1 · Najibi et al. · 2018 [cited by applicant]
US 20180285699A1 · Kolouri et al. · 2018 [cited by applicant]
US 20180373942A1 · Noda et al. · 2018 [cited by applicant]
US 20190132520A1 · Gupta et al. · 2019 [cited by applicant]
US 20190147220A1 · Mccormac et al. · 2019 [cited by applicant]
US 20190279044A1 · Vasisht et al. · 2019 [cited by applicant]
US 20190303725A1 · Gurvich et al. · 2019 [cited by applicant]
US 20190370613A1 · Arngren et al. · 2019 [cited by applicant]
US 20200014761A1 · Kawaai et al. · 2020 [cited by applicant]
US 20200019825A1 · Frei · 2020 [cited by examiner]
US 20200335208A1 · Talmor et al. · 2020 [cited by applicant]
US 20200372660A1 · Li et al. · 2020 [cited by applicant]
US 20200394458A1 · Yu · 2020 [cited by examiner]
US 20210035287A1 · Kim et al. · 2021 [cited by applicant]
US 20210056315A1 · Kale et al. · 2021 [cited by applicant]
US 20210150192A1 · Han · 2021 [cited by examiner]
US 20210150721A1 · Yohanandan et al. · 2021 [cited by applicant]
US 20210176239A1 · Rose · 2021 [cited by applicant]
US 20210262204A1 · Tafazoli Bilandi et al. · 2021 [cited by applicant]
US 20210269046A1 · Hashimoto · 2021 [cited by examiner]
US 20210406605A1 · Oleson et al. · 2021 [cited by applicant]
US 20210406606A1 · Oleson et al. · 2021 [cited by applicant]
US 20210406607A1 · Oleson et al. · 2021 [cited by applicant]
US 20220044066A1 · Amato · 2022 [cited by examiner]
US 20220405574A1 · Bathen · 2022 [cited by examiner]
US 20230079463A1 · Jonnada et al. · 2023 [cited by applicant]
US 20230117712A1 · Jaber et al. · 2023 [cited by applicant]
US 20230176430A1 · Nakao et al. · 2023 [cited by applicant]
US 20230326183A1 · Kah · 2023 [cited by examiner]
CN 110245659A · 2019 [cited by applicant]
JP 2010147696A · 2010 [cited by applicant]
JP 2013176101A · 2013 [cited by applicant]
JP 2019008460A · 2019 [cited by applicant]
JP 2019192082A · 2019 [cited by applicant]
WO 2007009876A1 · 2007 [cited by applicant]
WO 2018073121A1 · 2018 [cited by applicant]
WO WO2018173121A1 · 2018 [cited by applicant]
WO WO2018232378A1 · 2018 [cited by applicant]
WO WO2019176235A1 · 2019 [cited by applicant]
International Search Report and Written Opinion for International Application No. PCT/IB2021/053953, mailed Aug. 2, 2021. (14 pages). [cited by applicant]
Jiao, L. et al. ‘A Survey of Deep Learning-based Object Detection’, IEEE Access, vol. 7, Sep. 5, 2019, pp. 128837-128868. [cited by applicant]
Chao et al. 2018 “Fast object classification method based on saliency detection,” 2018 11th International Symposium on Computational Intelligence and Design, IEEE, 1:374-377. [cited by applicant]
Kavasidis et al. 2018 “A Saliency-based Convolutional Neural Network for Table and Chart Detection in Digitized Documents,” Arxiv.org, Cornell University Library, 13 pages. [cited by applicant]
Szegedy et al. 2015 “Going deeper with convolutions,” 2015 IEEE Conference on Computer Vision and Pattern Recognition, IEEE 9 pages. [cited by applicant]
Wang et al. 2018 “Video Salient Object Detection via Fully Convolutional Networks,” IEEE Transactions on Image Processing, 27(1):38-49. [cited by applicant]
Yohanandan et al. 2018 “Saliency Preservation in Low-Resolution Grayscale Images,” 237-254, XP047635328. [cited by applicant]
International Preliminary Report on Patentability for International Application No. PCT/IB2021/053953, mailed Nov. 17, 2022. (9 pages). [cited by applicant]
Supplemental European Search Report for European Application No. 21800158.4, mailed May 6, 2024. (12 pages). [cited by applicant]
Lu, B., et al., “Automating Deep Neural Network Model Selection for Edge Inference”, 2019 IEEE First International Conference on Cognitive Machine Intelligence (CogMI), 2019: pp. 184-193. https://api.semanticscholar.org… [cited by applicant]
De Lange, M., et al., “Unsupervised Model Personalization While Preserving Privacy and Scalability: An Open Problem,” 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 2020, p… [cited by applicant]
Luo, T., et al., “Distributed Anomaly Detection Using Autoencoder Neural Networks in WSN for IoT,” 2018 IEEE International Conference on Communications (ICC), Kansas City, MO, USA, 2018, pp. 1-6, doi: 10.1109/ICC.2018.8… [cited by applicant]
Reasons for Rejections for Japanese Application No. 2022-567287, mailed Aug. 15, 2024. (5 pages). [cited by applicant]
Shimura et al., “Efficient Object Detection for Railway Crossing Combining Image Classification and Object Detection,” IEICE Technical Report [online], The Institute of Electronics, Information and Communciation Enginee… [cited by applicant]
U.S. Appl. No. 17/314,465, filed May 7, 2021, Lars Oleson et al. [cited by applicant]
U.S. Appl. No. 17/473,467, filed Sep. 13, 2021, Lars Oleson et al. [cited by applicant]
U.S. Appl. No. 17/473,564, filed Sep. 13, 2021, Lars Oleson et al. [cited by applicant]