Building an Image Search Service from Scratch - Matthew Rubashkin, PhD - ODSC West 2018

Learn how to build an image search service from scratch in this 40-minute workshop. Explore techniques for creating representations of image and text data, and discover efficient methods for similarity search. Gain practical skills to develop a quick semantic search model, regardless of dataset size. Dive into computer vision approaches, advanced applications, classification, image captioning, and code generation. Follow along with hands-on coding examples and improve search results through various approaches. Conclude by running a Streamlit demo to showcase your newly acquired skills in building an image search service.

Intro
ON THE MENU
PINTEREST SEARCH
CV Approaches
Advanced Applications
CLASSIFICATION
IMAGE CAPTIONING
CODE GENERATION
Code - Part 1
Code - Part 2
First Approach
Second Approach
Third Approach
Image Embeddings
Code - Part 4
Improving Results
Code - Part 5
Code - Part 6 - Run the Streamlit Demo