PyCon Israel 2022

๐Ÿ‡ฎ๐Ÿ‡ฑ Computer Vision for The Poor: How to easily reduce Deep Computer Vision to shallow NLP
06-28, 16:30โ€“16:50 (Asia/Jerusalem), PyData

Being able to classify images is at the heart of many recommender systems. In this talk, we will share a simple trick to make the task of building an image classifier as easy as building a standard text classifier.

Building a task-specific image classification solution typically requires leveraging Computer Vision transfer learning techniques. It involves manipulating complex deep learning models, applying non trivial image preprocessing and using expensive hardware.
But what if you could leverage existing image meta-data annotations to classify our images?

In this talk we will share a simple trick to make the task of building an image classifier as easy as building a standard text classifier. This reduction simplifies preprocessing and training and it also dramatically reduces the required hardware & computation time. This reduction is made possible by leveraging ready-made computer vision APIs provided by the public cloud vendors. These APIs extract semantic textual labels from images that in turn can be used to build simple, shallow NLP classifiers.

This simple reduction has helped us deliver fast & cheap Python-based image classification models to production and is widely used in Outbrain products.

Session language โ€“


Target audience โ€“

Data Scientists

Other (target audience) โ€“

Product Managers, Data Analysts

Experienced, hands on, software and algorithms manager with unique multidisciplinary knowledge and proven leading skills. He has a wide experience in the design and implementation of machine learning, recommender systems, NLP, data mining, and optimization algorithms. He has been managing and building small-medium diverse engineering teams for over a decade and currently, he is Recommendations Data Science Manager, leading a diverse group of algorithm engineers in charge of the key KPIs of Outbrainโ€™s recommender system.

Hila has been processing, analyzing, and generating algorithms for the past decade. After earning her masters (summa cum laude) at BIU NLP lab and publishing at elite academic venues such as EMNLP, she began to research & develop algorithms that analyze call center calls as a senior researcher at NICE. She published 4 US patents and presented academic posters at various venues during that time. For the past two years, she has worked as DS Guild Master & algorithm engineer at Outbrain, where she works on large-scale super-fast algorithms for the native ads field. Hila also loves to teach and share her experience and has talked at various meetups and conferences