【Talk】Un-rectifying Non-linear Networks and Optimizations～Please register to come 免費報名參加～

Speaker：
Wen-Liang Hwang(黃文良 中央研究院資訊科學研究所 研究員)
Research Fellow, Institute of Information Science, Academia Sinica, Taiwan

Host：杭學鳴教授(交大電子系)、彭文孝教授(交大資工系)
Time：July 1, 2020 (Wednesday, 13:20pm~15:00pm)
地點：R329 , 3F Engineering Building 3rd, NCTU (交大工程三館三樓329室)
Organizer：科技部AI專案計畫 - 基於生成模型的視訊壓縮
Co-Organizer：科技部人工智慧普適研究中心
Contact window：(03)5131206，張采瑮(交大校內分機31206)，tsaili@g2.nctu.edu.tw

Registration website：https://ppt.cc/f4qf4x

Fee：free 免費報名參加

Notice 注意事項：
(1) Attendee are required to wear mask. 防疫期間，請參加者配戴口罩。
(2) 參加活動之成員，需同意將活動進行中的照片授權給主辦單位使用。
(3) 主辦單位有修改活動內容之權利。

Abstract: We consider deep neural networks with rectifier activations and max-pooling from a signal representation perspective. In this view, such representations mark the transition from using a single linear representation for all signals to utilizing a large collection of affine linear representations that are tailored to particular regions of the signal space. We propose a novel technique to “unrectify” the nonlinear activations into data-dependent linear equations and constraints, from which we derive explicit expressions for the affine linear operators, their domains and ranges in terms of the network parameters. Such analysis may facilitate understanding networks and promote further theoretical insight from both the signal processing and machine learning communities. (IEEE TSP vol. 68, 2020)