Quick Checks

Directory	Command (Alan's Web Services) HOST=104.52.63.66
head-poses	curl -X POST -F file=@head-poses/data/straight.png $HOST:7002
image	curl -X POST -F file=@image/data/code.png $HOST:7003
voice	curl -X POST -F file=@voice/data/terry.mp3 $HOST:7001

Setup

See Docker-Checkout.md for running a pre-configured 11GB docker image

Hardware	Recommended
CPU	1 core
RAM	3 GB
Disk	30 GB

Package	Ubuntu 24.04 or Ubuntu under WSL2 / AWS
core	sudo apt install ffmpeg python3-pip tesseract-ocr
base	pip3 install mediapipe tf-keras transformers torch openai-whisper pytesseract
server	pip3 install flask waitress

Reference Docs

Repo	Link
head pose	https://ai.google.dev/edge/mediapipe/solutions/vision/face_landmarker
image to text	https://tesseract-ocr.github.io/tessdoc/Installation.html
text to mood	https://huggingface.co/SamLowe/roberta-base-go_emotions
voice to text	https://github.com/openai/whisper