Easily reduce Deep Computer Vision to shallow NLP
summary
Shows how to turn image classification into a text-classification problem by calling cloud computer-vision APIs (public cloud vendor vision/labeling APIs) to extract semantic textual labels from images, then training lightweight Python NLP classifiers on those labels instead of fine-tuning deep vision models. Emphasizes that this reduces image preprocessing, hardware needs and training time compared with transfer-learning on deep CNNs, and describes how this approach is used in production at Outbrain.
AI-written summary · about
related talks
42mDeep Learning for Computer Vision
04mUsing Computer Vision to Understand Big Visual Data
40mHow to Implement a Production Ready Deep Learning Model in 12 Weeks
30mIntroduction to Computer Vision with OpenCV
44mBuilding smarter solutions with no expertise in machine learning
45mIntroduction To Deep Learning
33mBuilding end-to-end computer vision solutions from deep learning models
42mUse Machine Learning in your code, without being a ML expert