
# Advertising based on Age, Gender
### 1. Goal
Displaying ads which is suitable with customers based on their ages and genders
### 2. Requirements
* No internet (for privacy issue)
* Capturing faces of people are standing before camera
* Using a deep learning model to detect age, gender

* Displaying suitable ads on the screen (video mp4 files)


### 3.Hardware
* 1 Raspberry Pi 4 Model B 4GB RAM
https://www.raspberrypi.org/products/raspberry-pi-4-model-b/
* 1 Camera module
https://www.raspberrypi.org/products/camera-module-v2/
* 1 microSD card
* 1 charger
* Monitor for displaying: with resolution from customer's requirement
* Maximum power is 15W (comparing with a light buld ~ 40W)

### 4.Software
* OS: Raspian
* Programming language: Python, Shell Script
* Framework, tools: tensorflow, opencv, PyQT(for Graphic User Interface)
### 5.Limitation
* The system will run locally and it won't be double check by an online model. The accuracy is acceptable (customer's requirements)
* Customer will have to update ads inside the system by themselves
* The hardware (is being used - Raspberry Pi 4) satisfies enough(economic issue) for this task. There is small gap between switching videos (< 0.5s)
### 6.Extra Feature
Model can predict:
* Sentiments
* Emotions (happy, sad, ...)
* Style and clothes (T-shirt, Dress,...)
* Body language
*
### 7.Questions
* Are there any detail requirements? Could you describe more about the tasks?
* Format of input video (mp4)? For now I can choose the resolution to export to screen, so I just need HDMI connection (or VGA) with output screen
* The hardware I'm using is Ras Pi 4 4GB Ram, is that okay? Is there any specific systems?
* Do you need a GUI app to run or just need output from my python code?
* For now, I'm split age into zones: smaller 25, 25~35, 35~45, higher 45. For gender: male and female. Is that okay?
* About model, as I recommended before, you can use your images to test. If the results are good enough for you and you want to use it, then you can pay me.
* When would you like to start it?
* How many days would you like to spend for this project?