Cloud AI Services: What they are and how to use them
summary
Overview of concrete tools, limits and examples: Google Cloud offers prebuilt APIs for text, image and video (Natural Language, Vision, Video Intelligence) that return entities, sentiments, OCR, landmark matches (example: Vision API returns landmark + 85% confidence) and video label time ranges (Video Intelligence timestamps labels). Building custom models uses TensorFlow/Keras CNNs with training, evaluation, hyperparameter tuning and hosted prediction endpoints. AutoML automates splits, ensembling and deployment and can reach near-top Kaggle performance quickly (AutoML hits Kaggle plateau in ~1 hour). Deployment options include online, batch or export to Docker.
AI-written summary · about
chapters (19)
- 0:00Start
- 0:15Global Big Data Conference
- 3:30Cloud Al Builder Tools
- 4:21Use a pre-trained model to accomplish common ML tasks
- 8:24Multi-label classification
- 8:52Optical Character Recognition (OCR)
- 9:17Invoking the Vision API
- 12:16Scenario 2: what type of cat is this?
- 12:34Step 2: Gather & preprocess data
- 16:11Deploying the model into production
- 20:21Review dataset
- 20:59Deploy Model
- 21:39Movie Genre Classification
- 21:49Custom Sentiment
- 22:25Pretrained Entity Extraction
- 22:40Custom Entity Extraction
- 23:49Spotify Developer API
- 25:12The overall performance curve...
- 27:06Summary
related talks
42mUse Machine Learning in your code, without being a ML expert
59mMachine Intelligence at Google Scale
51mCloud AI: How to Get Started Injecting AI Into Your Applications
47mBuilding smarter apps with Machine Learning, from magic to reality
44mBuilding smarter solutions with no expertise in machine learning
57mBuilding smarter solutions with Machine Learning, from magic to reality
54mBuilding smarter apps with Machine Learning, from magic to reality
42mMachine Learning APIs, from Training [...]