應用數位影像偵測機能於工作現場即時輔助人員做影像辨識，乃為一種補強人員之智慧機能。為了提升辨識效率與品質，以往的方法需要大量的影像資料來訓練模組。但取得大量資料本質上就繁瑣耗時，且於註釋資料意涵的過程中又容易出錯因而影響整體表現。近年來，採用合成影像來補充影像資料之技術逐漸受到探討，也提出了許多類似的替代方案，其中一種調控方案即為領域隨機化。本研究中乃深入分析了幾種領域隨機化之策略，藉以釐清這些策略在物件影像辨識表現上之影響，而目標則以工業物件套用合成影像為應用範圍。影像參數中，則將對物件方位與彩現技術作深入探討。實作上則依研究所得之見解調控參數產出合成影像，套用於補足影像資料之辨識過程。本研究方法對於實境畫面相當欠缺之狀況下，提供了合成影像之調控策略藉以替入補足訓練資料，有效提升後續辨識之表現，實驗數據顯示當參數調控得宜可獲得高達37%之效能提升。

Real-time object detection can be a smart tool in assisting in-field human workers. In order to achieve a better performance out of object detection models, large quantities of data are needed to train the models’ network. However, obtaining this data can be costly and highly time-consuming, prone to errors in the annotating process which reduces the detection performance. Hence, in recent years, alternative ways to supplement the real data with synthetic data are continuously studied. One of the methods to generate them is called domain randomization. In this research, an extensive analysis of the impact of various domain randomization strategies was conducted to discover the impact on object detection among the strategies, so as to generate synthetic data from industrial objects. The strategies of object orientation and rendering randomization were comprehensively inspected. With this research, insights on the usage of better synthetic data generation are given to improve object detection performance in selected aspects. The method is especially efficient when extremely limited real data were available, and the delicate employment of synthetic data has strengthened the object detection performance up to 37% of improvement.

Abstract...i
摘要...ii
Acknowledgements...iii
Table of Contents...iv
List of Tables...vii
List of Figures...viii
Chapter 1. Introduction...1
1.1. Background...1
1.2. Motivation...2
1.3. Research Objective...4
1.4. Limitation...4
Chapter 2. Literature Review...5
2.1. Machine Learning...5
2.2. Convolutional Neural Network...5
2.3. Object Detection...7
2.3.1. YOLO...8
2.3.2. YOLOv4...9
2.3.3. YOLOv7...10
2.4. Metrics...12
2.4.1. Precision & Recall...12
2.4.2. Intersection over Union...12
2.4.3. mean Average Precision...13
2.5. Domain Randomization...13
2.6. Literature Summary...15
Chapter 3. Methodology...17
3.1. Dataset Generation...17
3.1.1. Real-World Image Data Generation...18
3.1.2. Synthetic Image Data Generation...22
3.2. Model Training & Evaluation...33
Chapter 4. Implementation Results & Discussion...36
4.1. Experiment Setup...36
4.2. Dataset Description...37
4.3. Baseline I & II...38
4.4. Object-Oriented Randomization...39
4.5. Environment / Rendering Randomization...39
4.6. Best Configuration...40
4.7. Comparison...42
Chapter 5. Conclusion & Future Work...45
5.1. Conclusion...45
5.2. Future Work...45
References...46
Appendix...49
1. Scale Randomizer...49
1.1. ScaleRandomizer...49
1.2. ScaleRandomizerTag...49
2. HDRI Sky Randomizer...49
2.1. HDRIRotationRandomizer...49
2.2. HDRIRotationRandomizerTag...50
2.3. HDRIStyleRandomizer...50
2.4. HDRIStyleRandomizerTag...51
3. Light Randomizer...51
3.1. LightRandomizer...51
3.2. LightRandomizerTag...51
4. Saturation Randomizer...52
4.1. SaturationRandomizer...52
4.2. SaturationRandomizerTag...52
5. Film Grain Randomizer...53
5.1. GrainRandomizer...53
5.2. GrainRandomizerTag...53
Extended Abstract...55

Albawi, S., Mohammed, T., & Al-Zawi, S. (2017). Understanding of a Convolutional Neural Network. International Conference on Engineering and Technology, 1-6.
Bochkovskiy, A., Wang, C.-Y., & Liao, M. (2020). YOLOv4: Optimal Speed and Accuracy of Object Detection. doi:10.48550/ARXIV.2004.10934
Borkman, S., Crespi, A., Dhakad, S., Ganguly, S., Hogins, J., Jhang, Y.-C. Parisi, P. (2021). Unity perception: Generate Synthetic Data for Computer Vision. arXiv preprint arXiv:2107.04259.
Dehban, A., Borrego, J., Figueiredo, R., Moreno, P., Bernardino, A., & Santos-Victor, J. (2019). The Impact of Domain Randomization on Object Detection: A Case Study on Parametric Shapes and Synthetic Textures. 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2593-2600.
Everingham, M., Van Gool, L., Williams, C., Winn, J., & Zisserman, A. (2010). The Pascal Visual Object Classes (VOC) Challenge. International Journal of Computer Vision, 88, 303-338.
Geiger, A., Lenz, P., & Urtasun, R. (2012). Are We Ready for Autonomous Driving? The KITTI Vision Benchmark Suite. 2012 IEEE Conference on Computer Vision and Pattern Recognition (pp. 3354-3361). IEEE.
He, K., Zhang, X., Ren, S., & Sun, J. (2015). Spatial PYramid Pooling in Deep Convolutional Networks for Visual Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 37(9), 1904-1916.
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 770-778).
Helm, J., Swiergosz, A., Haeberle, H., Karnuta, J., Schaffer, J., Krebs, V., Ramkumar, P. (2020). Machine Learning and Artificial Intelligence: Definitions, Applications, and Future Directions. Current Reviews in Musculoskeletal Medicine, 13(1), 69-76.
Li, F., & Xi, Q.-G. (2021). DefectNet: Toward Fast and Effective Defect Detection. IEEE Transactions on Instrumentation and Measurement, 70, 1-9.
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Zitnick, C. (2014). Microsoft coco: Common Objects in Context. European Conference on Computer Vision, 740-755.
Liu, S., Qi, L., Qin, H., Shi, J., & Jia, J. (2018). Path Aggregation Network for Instance Segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, (pp. 8759-8768).
McCormac, J., Handa, A., Leutenegger, S., & Davison, A. (2016). SceneNet RGB-D: 5M Photorealistic Images of Synthetic Indoor Trajectories with Ground Truth. arXiv 1612.05079.
Michaelis, C., Mitzkus, B., Geirhos, R., Rusak, E., Bringmann, O., Ecker, A., Brendel, W. (2019). Benchmarking Robustness in Object Detection: Autonomous Driving when Winter. arXiv preprint arXiv:1907.07484.
O'Shea, K., & Nash, R. (2015). Am Introduction to Convolutional Neural Networks. doi:10.48550/ARXIV.1511.08458
Redmon, J., Divvala, S., Girschick, R., & Farhadi, A. (2015). You Only Look Once: Unified, Real-Time Object Detection. doi:10.48550/ARXIV.1506.02640
Reinhard, E., Heidrich, W., Debevec, P., Pattanaik, S., Ward, G., & Myszkowski, K. (2010). High Dynamic Range Imaging: Acquisition, Displaym and Image-Based Lighting. Morgan Kaufmann.
Roboflow Inc. (2022). Roboflow.
Roboflow, Inc. (2020). Roboflow Annotate.
Technologies, U. (n.d.). Unity: High Definition Render Pipelines. Retrieved from https://docs.unity3d.com/Packages/com.unity.render-pipelines.high-definition\@14.0
Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., & Abbeel , P. (2017). Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World. 2017 IEEE/RSJ international conference on intelligent robots and systems, 23-30.
Tremblay, J., Prakash, A., Acuna, D., Brophy, M., Jampani, V., Anil, C., Birchfield, S. (2018). Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 1082-1090.
Wang, C.-Y., Bochkovskiy, A., & Mark, H.-Y. (2022). YOLOv7: Trainable Bag-Of-Freebies Sets New State-Of-The-Art for Real-Time Object Detectors. arXiv preprint arXiv:2207.02696.
Wang, C.-Y., Liao, M., Wu, Y.-H., Chen, P.-Y., Hsieh, J.-W., & Yeh, I.-H. (2020). CSPNet: A New Backbone That Can Enhance Learning Capability of CNN. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (pp. 390-391).
Wu, B., Wan, A., Iandola, F., Jin, P., & Keutzer, K. (2017). SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving. 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 446-454.
Zeng, N., Wu, P.-S., Wang, Z.-D., Li, H., Liu, W., & Liu, X. (2022). A Small-Sized Object Detection Oriented Multi-Scale Feature Fusion Approach with Application to Defect Detection. IEEE Transactions on Instrumentation and Measurement, 71, 1-14.
Zhao, H., Zhang, Y., Liu, S., Shi, J., Loy, C., Lin, D., & Jia, J. (2018). PSANet: Point-Wise Spatial Attention Network for Scene Parsing. Proceedings of the European Conference on Computer Vision, (pp. 267-283).
Zhao, Z.-Q., Zheng, P., Xu, S.-t., & Wu, X. (2019). Object Detection with Deep Learning: A Review. IEEE Transactions on Neural Networks and Learning Systems, 30(11), 3212-3232.