Applying Deep Learning in Content Based Information Retrieval System

Language

Chinese talk w. Chinese slides

Category

Science

Python Level

Intermediate

Abstract

隨著大數據時代的到來與機器學習的演算法逐漸的成熟和突破,市場上對於資料科學家的需求更是供不應求。其中,機器學習領域中的一支:深度學習(Deep Learning),近年在影像辨識、物件偵測、語音識別、自然語言處理、文章分類、情感分析、文字翻譯等領域有著相當顯著的進步。在日常生活中也隨處可以見到深度學習的應用,如Google的收尋引擎,手機的語音輸入與手寫辨識系統,Facebook上傳照片的自動標示等服務。同時,各種深度學習的Library也都傾巢而出,如Theano, Caffe, Torch和最新由Google推出的TensorFlow等。這顯示著深度學習漸漸走出學術領域,慢慢的走入市場上的應用與服務。本演講會針對深度學習進行深入淺出的介紹,同時比較在Python上不同的深度學習Library,並Demo我們利用這些Library建立的圖像辨識和CBIR圖像搜尋的系統。最後,我們會在簡介深度學習在圖像領域外是如何應用的。
*本演講的聽眾需具備基本的機器學習觀念。

Description

1. 介紹Deep Learning及常用的Library 2. Cotent Based Information Retrieval System - 介紹CBIR系統 - 如何導入Deep Learning - NLP的CBIR - Image的CBIR 3. 利用Python實做Deep Learning的CBIR系統 4. 結論與Q&A

Eddie Lin 林子耘

This is Eddie Lin. I started from physics to engineering, now working in the field of data science and machine learning, also trying to incorporate with music, art, design and social good in the long run. My interests are in computer vision, data science, deep learning and ubiquitous computing.

林志豪 (Chris Lin)

Here is my website: http://twchrislin.com/
This is Chris Lin. I am a physicist, data scientist, machine learning specialist. I get excited about data science, mathematical modeling, machine learning, and deep learning. I’d like to to understand things clearly, explain them well, and provide data-driven solution. On this website, I summarize my projects and keep track of what is changing in Deep Learning field.