IP Library › Granted Patent US 12,573,029
Granted Patent B2
US 12,573,029 · App. 16/858,219 · Granted Mar 10, 2026

Image annotation using one or more neural networks

Inventors: Holger Roth (Rockville, MD); Dong Yang (North Bethesda, MD); Daguang Xu (Potomac, MD); Vishwesh Nath (Nashville, TN)
Assignee: NVIDIA Corporation
G06T7/0012G06F40/169G06N3/08G06T7/11G06T2207/20104
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 12,573,029
App. No.
16/858,219
Granted
Mar 10, 2026
Kind
B2
Abstract

Apparatuses, systems, and techniques are presented to predict annotations for objects in images. In at least one embodiment, one or more annotations corresponding to one or more objects within one or more images are generated based, at least in part, on one or more neural networks iteratively trained using the one or more images.

Claims (79)

1 . One or more processors, comprising:

circuitry to:

predict, by using one or more neural networks and during a training iteration for the one or more neural networks, one or more objects in one or more labeled images of a training data set;

determine one or more values corresponding to one or more unlabeled images that indicate an uncertainty in predictions computed from the one or more unlabeled images;

determine one or more scores corresponding to the one or more unlabeled images based, at least in part, on the one or more values and mutual information between the one or more unlabeled images and the one or more labeled images;

select, based on the one or more scores, a set of unlabeled images from the one or more unlabeled images;

generate a label for the one or more selected unlabeled images to become one or more other labeled images;

add the one or more other labeled images to the training data set to increase uncertainty in predictions of a subsequent training iteration; and

update the one or more neural networks to predict the one or more objects by using the training data set including the added one or more other labeled images.

2 . The one or more processors of claim 1 , wherein the circuitry is to:

determine one or more intensity profiles of the one or more labeled images and the one or more unlabeled images; and

select, based at least on the one or more intensity profiles, the one or more unlabeled images.

3 . The one or more processors of claim 2 , wherein the one or more neural networks are trained in part by reusing images of the added one or more other labeled images for one or more training iterations based at least in part upon the uncertainty.

4 . The one or more processors of claim 2 , wherein the one or more neural networks are trained using a Stein variational gradient descent (SVGD) with a Dice log-likelihood.

5 . The one or more processors of claim 2 , wherein the circuitry is to select the one or more unlabeled images based at least on mutual information between one or more images of the set of unlabeled images and one or more images of a set of labeled images.

6 . The one or more processors of claim 1 , wherein the one or more unlabeled images comprise volumetric medical imaging data and the one or more objects are human organs.

7 . The one or more processors of claim 1 , wherein the one or more values comprise one or more entropy scores, and the one or more scores is based at least on subtracting the mutual information from the one or more values.

8 . A system comprising:

one or more processors to:

predict, by using one or more neural networks and during a training iteration for the one or more neural networks, one or more objects in one or more labeled images of a training data set;

determine one or more values corresponding to one or more unlabeled images that indicate an uncertainty in predictions computed from the one or more unlabeled images;

determine one or more scores corresponding to the one or more unlabeled images based, at least in part, on the one or more values and mutual information between the one or more unlabeled images and the one or more labeled images;

select, based on the one or more scores, a set of unlabeled images from the one or more unlabeled images;

generate a label for the selected unlabeled images to become one or more other labeled images;

add the one or more other labeled images to the training data set to increase uncertainty in predictions of a subsequent training iteration; and

update the one or more neural networks to predict the one or more objects by using the training data set including the added one or more other labeled images.

9 . The system of claim 8 , wherein the one or more processors are to:

determine one or more intensity profiles of the one or more labeled images and the one or more unlabeled images; and

select, based at least on the one or more intensity profiles, the one or more unlabeled images.

10 . The system of claim 9 , wherein the one or more neural networks are trained in part by reusing images of the added one or more other labeled images for one or more training iterations based at least in part upon the uncertainty.

11 . The system of claim 9 , wherein the one or more neural networks are trained using a Stein variational gradient descent (SVGD) with a Dice log-likelihood.

12 . The system of claim 9 , wherein the one or more processors are to select the one or more unlabeled images based at least on mutual information between one or more images of the set of unlabeled images and one or more images of a set of labeled images.

13 . The system of claim 8 , wherein the one or more unlabeled images comprise volumetric medical imaging data and the one or more objects are human organs.

14 . A method comprising:

predicting, by using one or more neural networks and during a training iteration for the one or more neural networks, one or more objects in one or more labeled images of a training data set;

determining one or more values corresponding to one or more unlabeled images that indicate an uncertainty in predictions computed from the one or more unlabeled images;

determining one or more scores corresponding to the one or more unlabeled images based, at least in part, on the one or more values and mutual information between the one or more unlabeled images and the one or more labeled images;

selecting, based on the one or more scores, a set of unlabeled images from the one or more unlabeled images;

generating a label for the selected unlabeled images to become one or more other labeled images;

adding the one or more other labeled images to the training data set to increase uncertainty in predictions of a subsequent training iteration; and

updating the one or more neural networks to predict the one or more objects by using the training data set including the added one or more other labeled images.

15 . The method of claim 14 , further comprising:

determining one or more intensity profiles of the one or more labeled images and the one or more unlabeled images; and

selecting, based at least on the one or more intensity profiles, the one or more unlabeled images.

16 . The method of claim 15 , wherein the one or more neural networks are trained in part by reusing images of the added one or more other labeled images for one or more training iterations based at least in part upon the uncertainty.

17 . The method of claim 15 , wherein the one or more neural networks are trained using a Stein variational gradient descent (SVGD) with a Dice log-likelihood.

18 . The method of claim 15 , further comprising selecting the one or more unlabeled images based at least on mutual information between one or more images of the set of unlabeled images and one or more images of a set of labeled images.

19 . The method of claim 14 , wherein the one or more unlabeled images comprise volumetric medical imaging data and the one or more objects are human organs.

20 . A machine-readable medium having stored thereon a set of instructions that, if performed by one or more processors, cause the one or more processors to at least:

predict, by using one or more neural networks and during a training iteration for the one or more neural networks, one or more objects in one or more labeled images of a training data set;

determine one or more values corresponding to one or more unlabeled images that indicate an uncertainty in predictions computed from the one or more unlabeled images;

determine one or more scores corresponding to the one or more unlabeled images based, at least in part, on the one or more values and mutual information between the one or more unlabeled images and the one or more labeled images;

select, based on the one or more scores, a set of unlabeled images from the one or more unlabeled images;

generate a label for the one or more selected unlabeled images to become one or more other labeled images;

add the one or more other labeled images to the training data set to increase uncertainty in predictions of a subsequent training iteration; and

update the one or more neural networks to predict the one or more objects by using the training data set including the added one or more other labeled images.

21 . The machine-readable medium of claim 20 , wherein the one or more processors are to:

determine one or more intensity profiles of the one or more labeled images and the one or more unlabeled images; and

select, based at least on the one or more intensity profiles, the one or more unlabeled images.

22 . The machine-readable medium of claim 21 , wherein the one or more neural networks are trained in part by reusing images of the added one or more other labeled images for one or more training iterations based at least in part upon the uncertainty.

23 . The machine-readable medium of claim 21 , wherein the one or more neural networks are trained using a Stein variational gradient descent (SVGD) with a Dice log-likelihood.

24 . The machine-readable medium of claim 21 , wherein the one or more processors are to select the one or more unlabeled images based at least on mutual information between one or more images of the set of unlabeled images and one or more images of a set of labeled images.

25 . The machine-readable medium of claim 20 , wherein the one or more unlabeled images comprise volumetric medical imaging data and the one or more objects are human organs.

26 . An image annotation system, comprising:

one or more processors to:

predict, by using one or more neural networks and during a training iteration for the one or more neural networks, one or more objects in one or more labeled images of a training data set;

determine one or more values corresponding to one or more unlabeled images that indicate an uncertainty in predictions computed from the one or more unlabeled images;

determine one or more scores corresponding to the one or more unlabeled images based, at least in part, on the one or more values and mutual information between the one or more unlabeled images and the one or more labeled images;

select, based on the one or more scores, a set of unlabeled images from the one or more unlabeled images;

generate a label for the selected unlabeled images to become one or more other labeled images;

add the one or more other labeled images to the training data set to increase uncertainty in predictions of a subsequent training iteration; and

update the one or more neural networks to predict the one or more objects by using the training data set including the added one or more other labeled images.

27 . The image annotation system of claim 26 , wherein the one or more processors are to:

determine one or more intensity profiles of the one or more labeled images and the one or more unlabeled images; and

select, based at least on the one or more intensity profiles, the one or more unlabeled images.

28 . The image annotation system of claim 27 , wherein the one or more neural networks are trained in part by reusing images of the added one or more other labeled images for one or more training iterations based at least in part upon the uncertainty.

29 . The image annotation system of claim 28 , wherein the one or more neural networks are trained using a Stein variational gradient descent (SVGD) with a Dice log-likelihood.

30 . The image annotation system of claim 28 , wherein the one or more processors are to select the one or more unlabeled images based at least on mutual information between one or more images of the set of unlabeled images and one or more images of a set of labeled images.

31 . The image annotation system of claim 27 , wherein the one or more unlabeled images comprise volumetric medical imaging data and the one or more objects are human organs.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 1, 2020
From: ROTH, HOLGER; YANG, DONG; XU, DAGUANG; NATH, VISHWESH
To: NVIDIA CORPORATION
Reel/Frame 052552/0001 →
Continuity (1)
Related Publication 20210334955A1 · Oct 28, 2021
References Cited (116)
US 10140544B1 · Zhao et al. · 2018 [cited by applicant]
US 10262236B2 · Lim · 2019 [cited by examiner]
US 10706542B2 · Chefd'hotel · 2020 [cited by examiner]
US 10885392B2 · Znamenskiy · 2021 [cited by examiner]
US 10885400B2 · Zhao · 2021 [cited by examiner]
US 10929708B2 · Syeda-Mahmood · 2021 [cited by examiner]
US 11093793B2 · Amato · 2021 [cited by examiner]
US 11100643B2 · Yang · 2021 [cited by examiner]
US 11105925B2 · Pacala · 2021 [cited by examiner]
US 11151721B2 · Avendi · 2021 [cited by examiner]
US 11188773B2 · Gao · 2021 [cited by examiner]
US 20150112182A1 · Sharma · 2015 [cited by examiner]
US 20170169567A1 · Chefd'hotel · 2017 [cited by examiner]
US 20170337682A1 · Liao · 2017 [cited by examiner]
US 20180276815A1 · Xu et al. · 2018 [cited by applicant]
US 20180322366A1 · Lim · 2018 [cited by examiner]
US 20190056497A1 · Pacala · 2019 [cited by examiner]
US 20190065901A1 · Amato · 2019 [cited by examiner]
US 20190304480A1 · Narayanan · 2019 [cited by examiner]
US 20190347524A1 · Znamenskiy · 2019 [cited by examiner]
US 20200012904A1 · Zhao et al. · 2020 [cited by applicant]
US 20200027561A1 · Rao · 2020 [cited by applicant]
US 20200167586A1 · Gao · 2020 [cited by examiner]
US 20200372635A1 · Veidman · 2020 [cited by examiner]
US 20210124999A1 · Dia · 2021 [cited by examiner]
US 20210334955A1 · Roth · 2021 [cited by examiner]
CA 2258941A1 · 1997 [cited by examiner]
CA 2425075A1 · 2002 [cited by applicant]
CN 107403446A · 2017 [cited by applicant]
CN 107492099A · 2017 [cited by applicant]
CN 108665462A · 2018 [cited by applicant]
CN 109074665A · 2018 [cited by applicant]
CN 109313717A · 2019 [cited by applicant]
CN 109478423A · 2019 [cited by applicant]
CN 110050281A · 2019 [cited by applicant]
CN 110176054A · 2019 [cited by applicant]
CN 110689037A · 2020 [cited by applicant]
CN 110689089A · 2020 [cited by applicant]
WO WO2019026081A2 · 2019 [cited by examiner]
WO 2020014778A1 · 2020 [cited by applicant]
WO 2020051053A1 · 2020 [cited by applicant]
Combined Search and Examination Report issued in GB Application No. 2105856.5, dated Oct. 18, 2021. [cited by applicant]
Information Processing in Medical Imaging, IPMI 2011, Springer, vol. 6801, pp. 25-36, Iglesias et al, “Combining Generative and Discriminative Models for Semantic Segmentation of CT Scans via Active Learning”, https://d… [cited by applicant]
IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2009, Joshi et al, “Multi-class active learning for image classification”, (Joshi). [cited by applicant]
arxiv.org, Gorriz et al, “Cost-Effective Active Learning for Melanoma Segmentation”, 2017, available from https://arxiv.org/pdf/1711.09168.pdf [accessed Oct. 14, 2021], (Gorriz). [cited by applicant]
arxiv.org, Wang et al, “Cost-Effective Active Learning for Deep Image Classification”, 2017, available from https://arxiv.org/pdf/1701.03551.pdf [accessed Oct. 14, 2021], (Wang). [cited by applicant]
arxiv.org, Yoo et al, “Learning Loss for Active Learning”, 2019, available from https://arxiv.org/pdf/1905.03677.pdf [accessed Oct. 14, 2021], (Yoo). [cited by applicant]
arxiv.org, Budd et al, “A Survey on Active Learning and Human-in-the-Loop Deep Learning for Medical Image Analysis”, 2019, available from https://arxiv.org/pdf/1910.02923vl.pdf [accessed Oct. 14, 2021], (Budd). [cited by applicant]
Angluin, “Queries and Concept Learning,” Machine learning 2(4): 1988, 24 pages. [cited by applicant]
Beluch et al., “The Power of Ensembles for Active Learning in Image Classification,” CVPR, 2018, 10 pages. [cited by applicant]
Bishop et al., “Neural Networks for Pattern Recognition,” Oxford University Press, 1995, 498 pages. [cited by applicant]
Cohn et al., “Active Learning with Statistical Models,” Journal of Artificial Intelligence Research, 1996, 8 pages. [cited by applicant]
Dolz et al., “Deep CNN Ensembles and Suggestive Annotations for Infant Brain MRI Segmentation,” Dec. 19, 2017, 20 pages. [cited by applicant]
Gal “Uncertainty in Deep Learning,” Ph.D. Dissertation, University of Cambridge, 2016, 174 pages. [cited by applicant]
Gal et al., “Deep Bayesian Active Learning with Image Data,” Proceedings of the 34th International Conference on Machine Learning—vol. 70, Mar. 8, 2017, 10 pages. [cited by applicant]
Gal et al., “Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning,” International Conference on Machine Learning, 2016, 11 pages. [cited by applicant]
Ghesu et al., “Quantifying and Leveraging Classification Uncertainty for Chest Radiograph Assessment,” vol. 11769, 2019, 9 pages. [cited by applicant]
He et al., “Deep Residual Learning for Image Recognition,” CVPR, 2016, 9 pages. [cited by applicant]
IEEE, “IEEE Standard 754-2008 (Revision of IEEE Standard 754-1985): IEEE Standard for Floating-Point Arithmetic,” Aug. 29, 2008, 70 pages. [cited by applicant]
Isensee et al., “nnU-Net: Breaking the Spell on Successful Medical Image Segmentation,” Apr. 17, 2019, 8 pages. [cited by applicant]
Isola et al., “Image-to-Image Translation with Conditional Adversarial Networks”, IEEE Conference on Computer Vision and Pattern Recognition, 2017, 10 pages. [cited by applicant]
Jungo et al., “Assessing Reliability and Challenges of Uncertainty Estimations for Medical Image Segmentation,” Oct. 11, 2019, 9 pages. [cited by applicant]
Kuo et al., “Cost-Sensitive Active Learning for Intracranial Hemorrhage Detection,” International Conference on Medical Image Computing and Computer-Assisted Intervention, Sep. 8, 2018, 8 pages. [cited by applicant]
Liu et al., “Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm,” Advances in Neural Information Processing Systems, 2016, 9 pages. [cited by applicant]
Ma et al., “Confidence Based Active Learning for Whole Object Image Segmentation,” International Workshop on Multimedia Content Representation, Classification and Security, 2006, 8 pages. [cited by applicant]
Mahapatra et al., “Efficient Active Learning for Image Classification and Segmentation using a Sample Selection and Conditional Generative Adversarial Network,” International Conference on Medical Image Computing and Co… [cited by applicant]
Milletari et al., “V-net: Fully convolutional neural networks for volumetric medical image segmentation,” 2016 Fourth International Conference on 3D Vision (3DV), Oct. 25, 2016, 11 pages. [cited by applicant]
Ozdemir et al., “Active Learning for Segmentation by Optimizing Content Information for Maximal Entropy,” Deep Learning in Medical Image Analysis andMultimodal Learning for Clinical Decision Support, 2018, 8 pages. [cited by applicant]
Ronneberger et al., “U-net: Convolutional networks for biomedical image segmentation,” International Conference on Medical Image Computing and Computer-Assisted Intervention, Oct. 5, 2015, 8 pages. [cited by applicant]
Settles, Active Learning, Synthesis Lectures on Artificial Intelligence and Machine Learning 6(1): Nov. 14, 2012, 68 pages. [cited by applicant]
Simpson et al., “A Large Annotated Medical Image Dataset for the Development and Evaluation of Segmentation Algorithms,” Feb. 25, 2019, 15 pages. [cited by applicant]
Sourati et al., “Active Deep Learning with Fisher Information for Patch-Wise Semantic Segmentation,” Deep Learning in Medical Image Analysis andMultimodal Learning for Clinical Decision Support, 2018, 11 pages. [cited by applicant]
Szegedy et al., “Going Deeper with Convolutions,” CVPR, 2015, 9 pages. [cited by applicant]
Tong et al., “Support Vector Machine Active Learning with Applications to Text Classification,” Journal of Machine Learning Research 2, 2001, 22 pages. [cited by applicant]
Top et al., “Active Learning for Interactive 3D Image Segmentation,” International Conference on Medical Image Computing and Computer-Assisted Intervention, 2011, 8 pages. [cited by applicant]
Wang et al., “Aleatoric Uncertainty Estimation with Test-Time Augmentation for Medical Image Segmentation with Convolutional Neural Networks,” Neurocomputing, 2019, 31 pages. [cited by applicant]
Xu et al., “Quantization of Fully Convolutional Networks for Accurate Biomedical Image Segmentation,” CVPR, 2018, 9 pages. [cited by applicant]
Yang et al., “Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentation,” International Conference on Medical Image Computing and Computer-Assisted Intervention, Jun. 15, 2017, 8 pages. [cited by applicant]
Yu et al., “Uncertainty-Aware Self-Ensembling Model for Semi-Supervised 3D Left Atrium Segmentation,” Jul. 16, 2019. [cited by applicant]
Zheng et al., “Biomedical Image Segmentation via Representative Annotation,” Thirty-Third AAAI Conference on Artificial Intelligence, 2019, 8 pages. [cited by applicant]
Zhou et al., “Fine-Tuning Convolutional Neural Networks for Biomedical Image Analysis: Actively and Incrementally,” CVPR, 2017, 12 pages. [cited by applicant]
Ash et al., “Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds,” Jun. 9, 2019, 21 pages. [cited by applicant]
Bengar et al., “Reducing Label Effort: Self-Supervised meets Active Learning,” IEEE, Oct. 11, 2021, 9 pages. [cited by applicant]
Bengar et al., “When Deep Learners Change Their Mind: Learning Dynamics for Active Learning?,” Jul. 30, 2021, 10 pages. [cited by applicant]
Chitta et al., “Large-Scale Visual Active Learning with Deep Probabilistic Ensembles,” Nov. 30, 2018, 10 pages. [cited by applicant]
Gal et al., “Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning,” Jun. 6, 2015, 10 pages. [cited by applicant]
He et al., “DiNTS: Differentiable Neural Network Topology Search for 3D Medical Image Segmentation,” Mar. 29, 2021, 10 pages. [cited by applicant]
Houlsby et al., “Cold-start Active Learning with Robust Ordinal Matrix Factorization,” Jun. 21, 2014, 9 pages. [cited by applicant]
IEEE “IEEE Standard for Floating-Point Arithmetric”, Microprocessor Standards Committee of the IEEE Computer Society, IEEE Std 754-2008, dated Jun. 12, 2008. [cited by applicant]
Isensee et al., “Automated Design of Deep Learning Methods for Biomedical Image Segmentation,” Apr. 2, 2020, 55 pages. [cited by applicant]
Kingma et al. “Adam: A Method for Stochastic Optimization,” arXiv:1412.6980, dated Dec. 22, 2014, 9 pages. [cited by applicant]
Lai et al., “Joint Semi-supervised and Active Learning for Segmentation of Gigapixel Pathology Images with Cost-Effective Labeling,” IEEE, Nov. 1, 2021, 22 pages. [cited by applicant]
Li et al., “Shape-aware Semi-supervised 3D Semantic Segmentation for Medical Images,” Jul. 21, 2020, 10 pages. [cited by applicant]
Nath et al., “Diminishing Uncertainty within the Training Pool: Active Learning for Biomedical Image Segmentation,” EEE, Dec. 29, 2020, 19 pages. [cited by applicant]
Nath et al., “The Power of Proxy Data and Proxy Networks for Hyper-Parameter Optimization in Medical Image Segmentation,” Jul. 12, 2021, 10 pages. [cited by applicant]
Nguyen et al., “How to Measure Uncertainty in Uncertainty Sampling for Active Learning,” Jun. 18, 2021, 35 pages. [cited by applicant]
Ouyang et al., “Self-supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation,” Jul. 20, 2020, 19 pages. [cited by applicant]
Bener et al., “Active Learning for Convolutional Neural Networks: A Core-Set Approach,” ICLR, 2018, 13 pages. [cited by applicant]
Siméoni et al., “Rethinking Deep Active Learning: Using Unlabeled Data at Model Training,” IEEE, Nov. 20, 2019, 8 pages. [cited by applicant]
Society of Automotive Engineers On-Road Automated Vehicle Standards Committee “Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles”, Standard No. J3016-201806, dated Jun. … [cited by applicant]
Society of Automotive Engineers On-Road Automated Vehicle Standards Committee, “Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles,” Standard No. J3016-201609, issued Jan… [cited by applicant]
Tang et al., “Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image Analysis,” Nov. 29, 2021, 16 pages. [cited by applicant]
Wang et al., “Annotation-Efficient Deep Learning for Automatic Medical Image Segmentation,” In, “Nature Communications,” Oct. 8, 2021, 13 pages. [cited by applicant]
Wang et al., “Semi-supervised Active Learning for Instance Segmentation via Scoring Predictions,” Dec. 9, 2020, 13 pages. [cited by applicant]
Wang et al., “Test-Time Augmentation With Uncertainty Estimation for Deep Learning-Based Medical Image Segmentation,” Jul. 19, 2018, 9 pages. [cited by applicant]
Xia et al., “3D Semi-Supervised Learning with Uncertainty-Aware Multi-View Co-Training,” Feb. 24, 2020, 10 pages. [cited by applicant]
Yuan et al., “Cold-start Active Learning through Self-supervised Language Modeling,” Oct. 22, 2020, 14 pages. [cited by applicant]
Hu et al., “Applying SVGD to Bayesian Neural Networks for Cyclical Time-Series Prediction and Inference,” Jan. 17, 2019, 5 pages. [cited by applicant]
Office Action for Chinese Application No. 202110443337.X, mailed Aug. 31, 2023, 17 pages. [cited by applicant]
United Kingdom Search and Examination Report for Application No. GB2105856.5, mailed Dec. 19, 2023, 6 pages. [cited by applicant]
Office Action for Chinese Application No. 202110443337.0, mailed Mar. 23, 2024, 10 pages. [cited by applicant]
Office Action for United Kingdom Application No. GB2105856.5, mailed Jul. 3, 2024, 10 pages. [cited by applicant]
Office Action for United Kingdom Application No. GB2105856.5, mailed Oct. 22, 2024, 4 pages. [cited by applicant]
Office Action for United Kingdom Application No. GB2105856.5, mailed Sep. 24, 2024, 6 pages. [cited by applicant]
Notice of Intent to Grant for United Kingdom Application No. GB2105856.5, mailed Nov. 14, 2024, 2 pages. [cited by applicant]
Office Action for Chinese Application No. 202110443337.X, mailed Oct. 30, 2024, 28 pages. [cited by applicant]