server

Task was to transmit image from camera on raspberry pi through web interface to the PC. PC is connected to raspberry through local network. Enable camera on your raspberry pi in raspi-conf. Reboot. Save the following source code to a file on your raspberry (let’s say streaming.py) do sudo pip3 install picamera run python3 streaming.py on the target machine (PC) go to http://<your raspberry ip or host name>:8000. output from the camera must be streaming in realtime....

TensorFlow Serving, sources - library for serving machine learning models. Written in C++ and Python. Server is in C++. Requires Bazel - Google’s build tool. Doesn’t work with python 3. Probably fast. TensorFlow: How to freeze a model and serve it with a python API Building a Machine Learning App with AWS Lambda (slides) Pipeline.io - End-to-End, Continuous Spark ML + Tensorflow AI Data Pipelines, Sources Interesting thread. They propose to use “saved_model_cli binary (in tools/), which you can feed a SavedModel, and pass input data via files....