---
tags: CV team
---

# 技術研討會時程表
2/22~3/5 各 track 準備研究內容，3/8 號之後`每週二`研討
| 週次 | <center>主題</center> | 負責組別 | 日期 |
|:---------: |:------------------------------- |:--------:|:----:|
| Week 1 | **CRNN + CTC**<br/>[paper] [An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition](https://arxiv.org/pdf/1507.05717.pdf)<br/>[blog] 上面那篇 paper 的[中文翻譯懶人包](https://www.itread01.com/content/1543061244.html)<br/> [code] [Belval/CRNN](https://github.com/Belval/CRNN)| track A | 3/09 |
| Week 2 | **Filter pruning:** <br/>[paper] [PRUNING FILTERS FOR EFFICIENT CONVNETS](https://arxiv.org/pdf/1608.08710.pdf) <br/>[blog] [paper的中文講解1](https://www.itread01.com/yqxyc.html)<br/>[paper的中文講解2](https://zhuanlan.zhihu.com/p/63779916)<br/> | track B | 3/16 |
| Week 3 | [【一】Viola & Jones 傳統人臉檢測器](https://web.iitd.ac.in/~sumeet/viola-cvpr-01.pdf)</br>[【二】Deep ID 人臉識別](http://mmlab.ie.cuhk.edu.hk/pdf/YiSun_CVPR14.pdf)</br>[【三】CVPR 2020 人臉防偽冠軍paper (近期個人研究)](https://arxiv.org/pdf/2006.16028v1.pdf)| track C | 3/23 |
| Week 4 | **CTC**<br/>[paper] [Connectionist Temporal Classification: Labelling Unsegmented Sequence Data with Recurrent Neural Networks](https://www.cs.toronto.edu/~graves/icml_2006.pdf)<br/>[code] [ctc source code](https://github.com/tensorflow/tensorflow/blob/v1.15.0/tensorflow/python/ops/ctc_ops.py#L47-L178)<br/>1. tf.nn.ctc_loss<br/>2. tf.nn.ctc_beam_search_decoder<br/>[blog] [宜昌老師的解說](https://www.ycc.idv.tw/crnn-ctc.html) | track A | 4/06 |
| Week 5 | [code] [Pruning Filters for Efficient Convnets](https://github.com/Eric-mingjie/rethinking-network-pruning/tree/master/imagenet/l1-norm-pruning) | track B | 4/13 |
| Week 6 | [【一】Cascade CNN 人臉檢測](https://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Li_A_Convolutional_Neural_2015_CVPR_paper.pdf)</br>[【二】MTCNN 人臉檢測](https://kpzhang93.github.io/MTCNN_face_detection_alignment/paper/spl.pdf)</br>[【三】FaceNet 人臉識別](https://arxiv.org/pdf/1503.03832.pdf) | track C | 4/20 |
| Week 7 | **Text Detection**<br>[paper] [Character Region Awareness for Text Detection](https://arxiv.org/abs/1904.01941)<br/>[inference code] [clovaai/CRAFT-pytorch](https://github.com/clovaai/CRAFT-pytorch) <br/> [training code 1] **better!**[brooklyn1900/CRAFT_pytorch](https://github.com/brooklyn1900/CRAFT_pytorch) <br/>[training code 2] **bad!**[backtime92/CRAFT-Reimplementation](https://github.com/backtime92/CRAFT-Reimplementation)| track A | 5/11 |
| Week 8 | **Channel pruning:** <br/>[paper] [Learning Efficient Convolutional Networks through Network Slimming](https://openaccess.thecvf.com/content_ICCV_2017/papers/Liu_Learning_Efficient_Convolutional_ICCV_2017_paper.pdf) <br/> [code] [Eric-mingjie/network-slimming](https://github.com/Eric-mingjie/network-slimming) | track B | 5/18 |
| Week 9 | [【一】Face R-CNN 人臉檢測](https://arxiv.org/pdf/1706.01061.pdf)</br>[【二】ArcFace 人臉識別 ](https://arxiv.org/pdf/1801.07698.pdf)| track C | 5/26 |
| Week 10 | **Attention Mechanism**<br/>[paper] [A Feasible Framework for Arbitrary-Shaped Scene Text Recognition](https://arxiv.org/pdf/1912.04561.pdf)<br/>[code] [zhang0jhon/AttentionOCR](https://github.com/zhang0jhon/AttentionOCR) | track A | 6/1 |
| Week 11 | break | | 6/8 |
| Week 12 | [paper] [The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks](https://arxiv.org/pdf/1803.03635.pdf)| track B | 6/15 |
| Week 13 |[【一】SSH 人臉檢測](https://arxiv.org/pdf/1708.03979.pdf)</br> [【二】RetinaFace 人臉檢測](https://arxiv.org/pdf/1905.00641.pdf)</br> | track C | 6/22|
| Week 14 | 【一】[text augmentation](https://arxiv.org/pdf/2003.06606.pdf)</br> 【二】[Learning to See Through Obstructions](https://arxiv.org/pdf/2004.01180v1.pdf) | 晉昇 +track A | 6/29 |
| Week 15 | break | | 7/6 |
| Week 16 | YOLOv4 模型壓縮分享 | track B | 7/13 |
| Week 17 |【一】[TableNet](https://arxiv.org/pdf/2001.01469.pdf)</br>【二】[Towards Universal Representation Learning](https://arxiv.org/pdf/2002.11841.pdf)| track C | 7/20 |
| Week 18 | 智金處月會 | | 7/27 |
| Week 19 | data augmentation 實作 | track A | 7/27 |
| Week 20 | break | | 8/10 |
| Week 21 | CRNN 模型壓縮分享 | track B | 8/19 |
| Week 22 |【一】[Vision Transformer](https://arxiv.org/pdf/2010.11929.pdf) | track C | 8/26 |
| Week 23 | OCR algorithms 解密大平台 | track A | 9/2 |
| Week 24 | break | | 9/9 |
| Week 25 | JN 策展暫停一次 | | 9/16 |
| Week 26 | ONNX / QUANTIZATION 技術研究 + 個人黑歷史分享 | track B | 9/23 |
| Week 27 | 貫中分享 | track C | 9/30 |
| Week 28 | [Robust Reflection Removal with Reflection-free Flash-only Cues](https://arxiv.org/pdf/2103.04273.pdf) | track C | 10/7 |
| Week 29 | 驗證碼大亂鬥(前篇) | track A | 10/14 |
| Week 30 | break | | 10/21 |
| Week 30 | [Distilling the Knowledge in a Neural Network](https://arxiv.org/pdf/1503.02531.pdf) | track B | 10/28 |
| Week 31 | break | | 11/4 |
| Week 32 | 忘記他報什麼ㄌ??? | 昱睿 | 11/11 |
| Week 33 | break | | 11/18 |
| Week 34 | Rotated Object Detection | 軒彤 | 11/25 |
| Week 35 | 存摺影像專案介紹 + 遇到困難點 | 立晟 | 12/02 |
| Week 36 | Rotated Object Detection | 晉昇 | 12/09 |
| Week 37 | break | | 12/16 |
| Week 38 | [DocReader: Bounding-Box Free Training of a Document Information Extraction Model](https://arxiv.org/pdf/2105.04313.pdf) | 弘林 | 12/23 |
| Week 39 | | 柏霖 | 12/30 |
| Week 40 | | 杰恩 | 01/06 |

> 當周分享的 track 需要做會議紀錄，形式待補

:pushpin: 輪值順序
軒彤、立晟、晉昇、弘林、柏霖、杰恩