Sim-to-Real Transfer for Object Detection in Aerial Inspections of Transmission Towers
Open accessAbstract
Training deep learning models for object detection usually requires a large amount of data, a condition that is not common for most real-world applications, especially in the context of aerial imagery. One possible solution is the use of simulators to generate synthetic data. For a good generalization, the model must be able to learn on the simulated data and perform correctly on the real data, process known as sim-to-real transfer. In this work, we analyze the generation of synthetic data to account for a data-scarce real-world scenario, which includes aerial imagery and object detection of transmission towers and their components. We evaluate the impact of image-to-image translation methods as domain adaptation techniques. In this analysis, we explore training strategies to mitigate the domain shift between synthetic and real data. According to our experimental results, the use of domain-adapted data through image-to-image translation could slightly improve the detection performance in real test data when compared to training with raw synthetic images only or with small datasets of real data, although it was noted through a visual analysis that objects with small bounding boxes, like clamp, anchoring clamp and ball link, could be distorted or vanished by the application of image-to-image translation methods. Additionally, when only a small subset of real data is available, training with both real and synthetic data at once led to better detection results, surpassing combinations of pre-training on synthetic and fine-tuning on real data.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Cites
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 74,248 · cited 1× · see citing works
“This also has been the metric reported on most papers to compare state-of-the-art object detection models, for example on [17], [18], [19], [20], [24], and [49].”
BackgroundFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 56,035 · see citing works
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803 · cited 2× · see citing works
“The YOLO network, for example, interprets object detection as a single regression problem with the pixels of the image as inputs and the bounding boxes coordinates and class probabilities as outputs, making it simpler and extremely fast [24].”
BackgroundMethodMicrosoft COCO: Common Objects in Context
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460 · cited 2× · see citing works
“Transfer learning and the fine-tuning of models trained initially on publicly large datasets such as VOC [28] and COCO [29] is often the preferred direction for practitioners.”
BackgroundMethodRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684 · cited 3× · see citing works
“Methods such as R-CNN [17] and its extensions, Fast R-CNN [18], Faster R-CNN [19] andMask R-CNN [20], are all two-stage models and have been largely employed as successful strategies for real-world applications [21], [22],”
BackgroundMethodHe, K., Gkioxari, G., Dollar, P., & Girshick, R. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.322 · Cited by 29,137 · see citing works
Girshick, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.169 · Cited by 28,105 · see citing works
Unpaired image-to-image translation using cycle-consistent adversarial networks
Zhu, J.-Y., Park, T., Isola, P., & Efros, A. A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.244 · Cited by 23,602 · cited 3× · see citing works
“The following techniques were evaluated: CycleGAN [47] (Study A2): this method introduced the cycle consistency loss to regularize the mapping from one domain to another and back again.”
BackgroundMethodThe pascal visual object classes (VOC) challenge
Everingham, M., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2010) · International Journal of Computer Vision · doi:10.1007/s11263-009-0275-4 · Cited by 19,599 · see citing works
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
Zhang, R., Isola, P., Efros, A. A., Shechtman, E., & Wang, O. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00068 · Cited by 14,623 · cited 4× · see citing works
“Perceptual Image Patch Similarity (LPIPS) [48] to select the frames of interest.”
BackgroundInfluentialThe Cityscapes Dataset for Semantic Urban Scene Understanding
Cordts, M., Omran, M., Ramos, S., Rehfeld, T., Enzweiler, M., Benenson, R., Franke, U., Roth, S., & Schiele, B. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.350 · Cited by 12,433 · cited 1× · see citing works
“Real and simulation driving scenes from CityScapes [30] and GTAV [31] datasets.”
MethodFCOS: Fully Convolutional One-Stage Object Detection
Tian, Z., Shen, C., Chen, H., & He, T. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00972 · Cited by 7,013 · cited 2× · see citing works
“However, new one-stage detectors, such as FCOS [25] and YOLOv5 [26], have shown that one-stage solutions are also suitable for performance-dependent tasks.”
Representation Learning with Contrastive Predictive Coding
Oord, A. V. D., Li, Y., & Vinyals, O. (2018) · arXiv · doi:10.48550/arxiv.1807.03748 · Cited by 6,242 · see citing works
Object Detection With Deep Learning: A Review
Zhao, Z.-Q., Zheng, P., Xu, S.-T., & Wu, X. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2876865 · Cited by 5,545 · see citing works
CARLA: An open urban driving simulator
Dosovitskiy, A., Ros, G., Codevilla, F., Lopez, A., & Koltun, V. (2017) · arXiv · doi:10.48550/arxiv.1711.03938 · Cited by 3,718 · cited 1× · see citing works
“For the purpose of training and validating autonomous cars systems, the CARLA simulator was developed by Dosovitskiy et al. [37], which included the generation of semantic segmentation, object detection, depth, and LiDAR (Light Detection and Ranging) data.”
MethodObject Detection in 20 Years: A Survey
Zou, Z., Chen, K., Shi, Z., Guo, Y., & Ye, J. (2023) · Proceedings of the IEEE · doi:10.1109/jproc.2023.3238524 · Cited by 3,689 · cited 1× · see citing works
“Traditional techniques have explored the use of feature descriptors and other handcrafted features [13] as reference for object localization.”
BackgroundDomain randomization for transferring deep neural networks from simulation to the real world
Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., & Abbeel, P. (2017) · 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8202133 · Cited by 3,124 · cited 1× · see citing works
“Tobin et al. [40] explored the concept of domain randomization by training an object detection model on geometric objects with non-realistic random textures and demonstrated the model trained on the simulated data has better generalization when tested in the real domain.”
BackgroundCyCADA: Cycle-Consistent Adversarial Domain Adaptation
Hoffman, J., Tzeng, E., Park, T., Zhu, J.-Y., Isola, P., Saenko, K., Efros, A. A., & Darrell, T. (2017) · arXiv · doi:10.48550/arxiv.1711.03213 · Cited by 2,656 · cited 2× · see citing works
“…explored the application of style transfer and domain adaption techniques (techniques that aim to improve the performance of a model trained on data that belongs to a different domain) to compensate for these differences and improve results for applications in the real-world [8], [9], [10], [11].”
BackgroundMethodRos, G., Sellart, L., Materzynska, J., Vazquez, D., & Lopez, A. M. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.352 · Cited by 2,338 · cited 1× · see citing works
“between the generated and real data [6], [7].”
BackgroundAirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles
Shah, S., Dey, D., Lovett, C., & Kapoor, A. (2018) · Springer Proceedings in Advanced Robotics · doi:10.1007/978-3-319-67361-5_40 · Cited by 2,254 · cited 4× · see citing works
“Reducing costs and increasing time efficiency are the main factors since a limited effort of setting up a realistic simulation or a generative model can be worth a large volume of high-quality data [5].”
BackgroundMethodInfluentialEfficient Non-Maximum Suppression
Neubeck, A., & Van Gool, L. (2006) · 18th International Conference on Pattern Recognition (ICPR'06) · doi:10.1109/icpr.2006.479 · Cited by 2,118 · cited 1× · see citing works
“Two-stage detectors usually work by first identifying regions of interest or possible bounding boxes and then running a classifier on those regions with sometimes a post-processing step for filtering out duplicated detections [16].”
BackgroundPlaying for Data: Ground Truth from Computer Games
Richter, S. R., Vineet, V., Roth, S., & Koltun, V. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-46475-6_7 · Cited by 1,928 · see citing works
Learning from Simulated and Unsupervised Images through Adversarial Training
Shrivastava, A., Pfister, T., Tuzel, O., Susskind, J., Wang, W., & Webb, R. (2017) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2017.241 · Cited by 1,831 · cited 2× · see citing works
“…explored the application of style transfer and domain adaption techniques (techniques that aim to improve the performance of a model trained on data that belongs to a different domain) to compensate for these differences and improve results for applications in the real-world [8], [9], [10], [11].”
BackgroundDomain Adaptive Faster R-CNN for Object Detection in the Wild
Chen, Y., Li, W., Sakaridis, C., Dai, D., & Van Gool, L. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00352 · Cited by 1,672 · cited 2× · see citing works
“The distribution of the visual features present in the synthetic data domain may differ from those in the real domain distribution [10] as texture and light reflectance are hard to be represented computationally.”
BackgroundA Survey on Performance Metrics for Object-Detection Algorithms
Padilla, R., Netto, S. L., & da Silva, E. A. B. (2020) · 2020 International Conference on Systems, Signals and Image Processing (IWSSIP) · doi:10.1109/iwssip48289.2020.9145130 · Cited by 1,605 · cited 1× · see citing works
“95] [27].”
Contrastive Learning for Unpaired Image-to-Image Translation
Park, T., Efros, A. A., Zhang, R., & Zhu, J.-Y. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58545-7_19 · Cited by 1,605 · cited 2× · see citing works
“The hyperparameters described below are relative to the execution script of the official implementations [47], [51] and [11] of all three methods.”
BackgroundMethodUnsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
Bousmalis, K., Silberman, N., Dohan, D., Erhan, D., & Krishnan, D. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.18 · Cited by 1,603 · cited 1× · see citing works
“Bousmalis et al. [44] used a GAN approach on the Synthetic Cropped LineMod [45] dataset to make the synthetic 3D models closer to the target domain for object classification and position estimation.”
MethodRecent advances in deep learning for object detection
Wu, X., Sahoo, D., & Hoi, S. C. H. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.01.085 · Cited by 1,047 · cited 1× · see citing works
“However, with the recent advances on deep learning and the improvement of hardware capabilities, most of the state-of-the-art methods in this field has been built on top of deep neural networks [14], [15].”
BackgroundTraining Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization
Tremblay, J., Prakash, A., Acuna, D., Brophy, M., Jampani, V., Anil, C., To, T., Cameracci, E., Boochoon, S., & Birchfield, S. T. (2018) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2018.00143 · Cited by 936 · cited 1× · see citing works
“Tremblay et al. [41] applied domain randomization on object detection of cars by generating scenes with random position, rotation, and textures of the vehicles, as also including in the scene random geometric shapes to be used as negative examples and randomized lighting and position of the camera.”
BackgroundFace Detection with the Faster R-CNN
Jiang, H., & Miller, E. G. (2017) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg.2017.82 · Cited by 823 · cited 1× · see citing works
“Methods such as R-CNN [17] and its extensions, Fast R-CNN [18], Faster R-CNN [19] and Mask R-CNN [20], are all two-stage models and have been largely employed as successful strategies for real-world applications [21], [22], [23].”
MethodTao, X., Zhang, D., Wang, Z., Liu, X., Zhang, H., & Xu, D. (2018) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2018.2871750 · Cited by 793 · cited 1× · see citing works
“This is the case of object detection in aerial scenarios [2], [3], [4].”
BackgroundTartanAir: A Dataset to Push the Limits of Visual SLAM
Wang, W., Zhu, D., Wang, X., Hu, Y., Qiu, Y., Wang, C., Hu, Y., Kapoor, A., & Scherer, S. (2020) · 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros45743.2020.9341801 · Cited by 456 · cited 1× · see citing works
“With this tool, large datasets such as the Mid-Air [38] and TartanAir [39] benchmarks were created to advance research in robot navigation and visual understanding.”
MethodLearning descriptors for object recognition and 3D pose estimation
Wohlhart, P., & Lepetit, V. (2015) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2015.7298930 · Cited by 452 · cited 2× · see citing works
“Bousmalis et al. [44] used a GAN approach on the Synthetic Cropped LineMod [45] dataset to make the synthetic 3D models closer to the target domain for object classification and position estimation.”
MethodSynthetic Data for Deep Learning
Nikolenko, S. (2021) · Springer optimization and its applications · doi:10.1007/978-3-030-75178-4 · Cited by 434 · see citing works
A cost effective solution for pavement crack inspection using cameras and deep neural networks
Mei, Q., & Gül, M. (2020) · Construction and Building Materials · doi:10.1016/j.conbuildmat.2020.119397 · Cited by 253 · see citing works
Applying Faster R-CNN for Object Detection on Malaria Images
Hung, J., & Carpenter, A. E. (2017) · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2017.112 · Cited by 245 · cited 1× · see citing works
“Methods such as R-CNN [17] and its extensions, Fast R-CNN [18], Faster R-CNN [19] and Mask R-CNN [20], are all two-stage models and have been largely employed as successful strategies for real-world applications [21], [22], [23].”
MethodA closer look at Faster R-CNN for vehicle detection
Fan, Q., Brown, L., & Smith, J. V. (2016) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2016.7535375 · Cited by 219 · cited 1× · see citing works
“Methods such as R-CNN [17] and its extensions, Fast R-CNN [18], Faster R-CNN [19] and Mask R-CNN [20], are all two-stage models and have been largely employed as successful strategies for real-world applications [21], [22], [23].”
MethodTTPLA: An Aerial-Image Dataset for Detection and Segmentation of Transmission Towers and Power Lines
Abdelfattah, R., Wang, X., & Wang, S. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-69544-6_36 · Cited by 135 · cited 1× · see citing works
“This is the case of object detection in aerial scenarios [2], [3], [4].”
BackgroundMOTSynth: How Can Synthetic Data Help Pedestrian Detection and Tracking?
Fabbri, M., Brasó, G., Maugeri, G., Cetintas, O., Gasparini, R., Ošep, A., Calderara, S., Leal-Taixé, L., & Cucchiara, R. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01067 · Cited by 131 · cited 1× · see citing works
“between the generated and real data [6], [7].”
BackgroundTSIT: A Simple and Versatile Framework for Image-to-Image Translation
Jiang, L., Zhang, C., Huang, M.-Y., Liu, C., Shi, J., & Loy, C. C. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58580-8_13 · Cited by 129 · cited 3× · see citing works
“Two-Stream Image-to-image Translation - TSIT [11] (Study A4): this approach uses a two-stream network dedicated to integrating content and style information into the generated image.”
BackgroundMethodDetection and Evaluation Method of Transmission Line Defects Based on Deep Learning
Liang, H., Zuo, C., & Wei, W. (2020) · IEEE Access · doi:10.1109/access.2020.2974798 · Cited by 121 · see citing works
Mid-Air: A Multi-Modal Dataset for Extremely Low Altitude Drone Flights
Fonder, M., & Droogenbroeck, M. V. (2019) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2019.00081 · Cited by 115 · cited 1× · see citing works
“With this tool, large datasets such as the Mid-Air [38] and TartanAir [39] benchmarks were created to advance research in robot navigation and visual understanding.”
MethodHwang, H., Jang, C., Park, G., Cho, J., & Kim, I.-J. (2020) · IEEE Access · doi:10.1109/access.2021.3051842 · Cited by 89 · cited 1× · see citing works
“In this context, Hwang et al [34] used the Unreal Engine to develop ElderSim, a platform for generating synthetic data for human action recognition of elderly people.”
MethodBian, J., Hui, X., Zhao, X., & Tan, M. H. (2019) · International Journal of Advanced Robotic Systems · doi:10.1177/1729881418820227 · Cited by 43 · see citing works
GeneSIS-Rt: Generating Synthetic Images for Training Secondary Real-World Tasks
Stein, G. J., & Roy, N. (2018) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2018.8462971 · Cited by 25 · see citing works
Unlimited Road-scene Synthetic Annotation (URSA) Dataset
Angus, M., ElBalkini, M., Khan, S., Harakeh, A., Andrienko, O., Reading, C., Waslander, S. L., & Czarnecki, K. (2018) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc.2018.8569519 · Cited by 23 · cited 1× · see citing works
“Also using the GTAV environment, Angus et al. [36] created the URSA 110314 VOLUME 11, 2023 Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
MethodTang, H., & Jia, K. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01531 · Cited by 20 · cited 1× · see citing works
“Tang et al. [42] explored the usage of synthetic data generated with Blender [43] and studied a handful of aspects regarding the usage of such data for downstream segmentation tasks.”
Hu, Y.-T., Wang, J., Yeh, R. A., & Schwing, A. G. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00375 · Cited by 8 · cited 1× · see citing works
“Hu et al. [35] used the engine of the game Grand Theft Auto V (GTAV) to create a dataset called SAIL-VOS 3D for object detection and 3D mesh reconstruction from videos.”
Method
Flagged first, then by citations
1 of the 49 references couldn’t be matched to a known paper.
Research works printed but not linked: 4
- 2023, Unreal Engine
- 2018, Blender—A 3D Modelling and Rendering Package
- goodfellow, 2016, Deep Learning
- jocher, 2023, ultralytics yolov5
Cited by
Sorted by citations
Backman, K., Wood, J., Brandimarti, M. E., Beranek, C. T., & Roff, A. (2025) · Methods in Ecology and Evolution · doi:10.1111/2041-210x.70006 · Cited by 11 · see citing works
“Prior works have shown success in utilising AirSim to train deep learning models in simulation whilst being transferable to real - ity (Backman et al., 2023 ; Bonatti et al., 2020 ; Bondi et al., 2018 ; Peterlevitz et al., 2023 ) where (Bondi et al., 2018)”
Analyzing the Impact of Training Maritime UAV Object Detection Models on Synthetic and Real Data
Miller, D., & Lu, K. W. (2025) · doi:10.1109/wocc63563.2025.11082215 · Cited by 1 · see citing works
Freire, D. L., de Carvalho, A. C. P. L. F., Peterlevitz, A. J., et al. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73500-4_16 · Cited by 1 · see citing works
Self-citationCenter-Guided Network with Dynamic Attention for Transmission Tower Detection
Li, X., Liang, Z., Yang, J., Lyu, C., & Xu, Y. (2025) · Information · doi:10.3390/info16040331
Trescher, D., Haag, W., & Schröder, E. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-02813-6_22
Density-Based Counting in Cluttered Scenes via Synthetic Data and Deep Vision Models
Sevilmis, B., Onle, B., & Ozkan, O. (2025) · 2025 10th International Conference on Computer Science and Engineering (UBMK) · doi:10.1109/ubmk67458.2025.11206803
“…with (cid:1) a (cid:1) small (cid:1) real (cid:1) subset (cid:1) markedly boosts (cid:1) accuracy; (cid:1) as (cid:1) shown (cid:1) in (cid:1) [20], (cid:1) training (cid:1) with (cid:1) mixed synthetic+real (cid:1) data (cid:1) improves (cid:1) sim-to-real (cid:1) object (cid:1) detection…”
Liu, Y., Ni, Y., Wang, H., et al. (2026) · Advances in Structural Engineering · doi:10.1177/13694332261483422
- Authors
- Peterlevitz, A. J., Chinelatto, M. A., Menezes, A. G., Motta, C. A. M., Pereira, G. A. B., Lopes, G. L., Souza, G. D. M., Rodrigues, J., Godoy, L. C., Koller, M. A. F. F., Cabral, M. O., Alves, N. E., Silva, P. H., Cherobin, R., Yamamoto, R. A. O., & Silva, R. D. D.
- Affiliations
- Instituto de Pesquisas Eldorado (BR)
- Universidade de São Paulo (BR)
- Computer Vision Department, Eldorado Research Institute, Campinas, São Paulo, Brazil
- ORCID
- Venue
- IEEE Access
- Published
- 2023
- Volume
- 11
- Pages
- 110312-110327
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W4387385679
- Semantic Scholar
- e01f6ed17e4477c1c50c13c1097a9d1795385f38
- Cited by
- 7
- References
- 49
- Citations by year
- 2026: 1 · 2025: 5 · 2024: 1
- FWCI
- 0.63
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- IEEE Access (publishedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Transmissora Aliança de Energia Elétrica S.A. (TAESA) through Agência Nacional de Energia Elétrica (ANEEL) Research and Development Program (PD-07130-0059/2020)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Engineering › Mechanical Engineering › Power Line Inspection Robots
- Keywords
- sim-to-real transfer, object detection, aerial imagery, transmission towers, synthetic data generation, domain adaptation, image-to-image translation, domain shift, data scarcity, aerial inspection, pre-training and fine-tuning
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
