臺灣博碩士論文加值系統

摘要 i
ABSTRACT ii
目錄 iii
第一章 序論 1
1.1研究動機 1
1.2論文架構 1
第二章 文獻探討 2
2.1目標偵測（Object detection） 2
2.1.1 Two-stage: R-CNN 3
2.1.2 Two-stage: Fast R-CNN 3
2.1.3 Two-stage: Faster R-CNN 4
2.1.4 One-stage: YOLO v1 5
2.1.5 One-stage: YOLO v2 7
2.1.6 Fast-NMS 8
2.2雙階段實例分割（Two-stage Instance Segmentation） 8
2.2.1 Mask R-CNN 9
2.2.2 Mask Scoring R-CNN 10
2.3單階段實例分割（One-stage Instance Segmentation） 11
2.3.1 Yolact 11
2.3.2 Yolact ++ 14
2.3.3 YolactEdge 15
2.4光流估算-IFRNet 16
第三章 研究方法 19
3.1系統流程 19
3.2系統架構 21
3.2.1 Instance segmentation module 21
3.2.2 Association module 25
3.2.3 Interpolation module 28
第四章 實驗結果 30
4.1實驗環境 30
4.2實驗資料集 30
4.3評估方法 33
4.4實驗結果 33
4.4.1 YouTube-VIS 33
4.4.2 COCO 2017 Dataset 38
第五章 結論 40
第六章 參考文獻 41

[1]R. Girshick, J. Donahue, T. Darrell, J. Malik, “Rich feature hierarchies for accurate object detection and semantic segmentation,” CVPR, 2014.
[2]R. Girshick, “Fast R-CNN,” ICCV, 2015.
[3]S. Ren, K. He, R. Girshick, J. Sun, “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks,” arXiv:1506.01497, 2015.
[4]K. He, G. Gkioxari, P. Dollár, R. Girshick, “Mask R-CNN,” arXiv:1703.06870, 2015.
[5]J. Redmon, S. Divvala, R. Girshick, A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” arXiv:1506.02640, 2015.
[6]J. Redmon, A. Farhadi, “YOLO9000: Better, Faster, Stronger,” arXiv:1612.08242, 2016.
[7]D. Bolya, C. Zhou, F. Xiao, Y. Jae Lee, “YOLACT: Real-time Instance Segmentation,” ICCV, 2019.
[8]Tsung-Yi Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, S. Belongie, “Feature Pyramid Networks for Object Detection,” arXiv:1612.03144, 2016.
[9]W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, Cheng-Yang Fu, A. C. Berg, “SSD: Single Shot MultiBox Detector,” ECCV, 2016.
[10]J. Long, E. Shelhamer, T. Darrell, “Fully Convolutional Networks for Semantic Segmentation,” CVPR, 2015.
[11]L. Kong, B. Jiang, D. Luo, W. Chu, X. Huang, Y. Tai, C. Wang, J. Yang, “IFRNet: Intermediate Feature Refine Network for Efficient Frame Interpolation,” CVPR, 2022.
[12]A. Bewley, Z. Ge, L. Ott, F. Ramos, B. Upcroft, “Simple Online and Realtime Tracking,” ICIP, 2016.
[13]Tsung-Yi Lin, M. Maire, S. Belongie, L. Bourdev, R. Girshick, J. Hays, P. Perona, D. Ramanan, C. Lawrence Zitnick, P. Dollár, “Microsoft COCO: Common Objects in Context,” arXiv:1405.0312, 2014.
[14]N. Xu, L. Yang, Y. Fan, D. Yue, Y. Liang, J. Yang, T. Huang, “YouTube-VOS: A Large-Scale Video Object Segmentation Benchmark,” arXiv:1809.03327, 2018.
[15]Y. Tang, “Deep Learning using Linear Support Vector Machines,” ICML, 2013.
[16]C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, A. Rabinovich, “Going Deeper with Convolutions,” arXiv:1409.4842, 2014.
[17]J. Redmon, A. Farhadi, “YOLOv3: An Incremental Improvement” arXiv:1804.02767, 2018.
[18]Tsung-Yi Lin, P. Goyal, R. Girshick, K. He, P. Dollár, “Focal Loss for Dense Object Detection,” arXiv:1708.02002, 2017.
[19]M. Tan, Q. V. Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” ICML, 2019.
[20]A. Bochkovskiy, Chien-Yao Wang, Hong-Yuan Mark Liao, “YOLOv4: Optimal Speed and Accuracy of Object Detection,” arXiv:2004.10934, 2020.
[21]J. Dai, H. Qi, Y. Xiong, Y. Li, G. Zhang, H. Hu, Y. Wei, “Deformable Convolutional Networks,” ICCV, 2017.
[22]Z. Huang, L. Huang, Y. Gong, C. Huang, X. Wang, “Mask Scoring R-CNN,” CVPR, 2019.
[23]D. Bolya, C. Zhou, F. Xiao, Y. Jae Lee, “YOLACT++: Better Real-time Instance Segmentation,” arXiv:1912.06218, 2019.
[24]H. Liu, R. A. Rivera Soto, F. Xiao, Y. Jae Lee, “YolactEdge: Real-time Instance Segmentation on the Edge,” arXiv:1912.06218, 2020.
[25]K. He, X. Zhang, S. Ren,, and J. Sun, “Deep residual learning for image recognition.” CVPR, pp. 770-778, 2016.