Gemini-AI-lens-2025- 💛✨🌹

Gemini API Real-Time, future glasses with built-in AI 🚀 **Large project using Gemini RAG, This project was made with an original idea "Make an AI glasses with built in AI", And yes, I am getting this, so please leave a star

NOTE: I was 13 years old in 2025, when I made this project

Note Subject	Note Description	Year Wrote
Motivation	Valentina Werle (Werle) motivated the project and the addition of psychologist mode. (she and a friend told me the idea) 🌹	2025
Project Timeline	The project took 3 months of coding + 1 month of investigations. ✨	2025
Development	The project was developed alone, without any human help. 😉	2025
Language	The author's native language is Spanish, but they also dominate English. (Written for future reference on English level in 2025) 🧇	2025
Priorities	I'am prioritized low-cost + efficiency over ease of software. 🔱	2025
Contribution	I'am open to contributions to Open Source projects. 💫	2025
Time Spent	4 Month's Aproximated Time 💦	2025

---

📌 Características

• 🌐 With Built In AI vision
• ⚡ Realtime proccesing
• 🛠️ Omptimized for low-cost hardware
• 🎯 Easy management UI for easy usage

---

3 Different Modes;

1. Psicologo ⚡ PRE-BETA
2. Conversacional 😊 PRE-BETA
3. Lentes (Voice Activation) ✨ SOON

---

🔍 Workflow's

Here is some workflow's that I made just for let you know how it work 😁

🛠️ Fully Native Memory SYS

⚡ Reconnect Rate Limit (2 Mins limit to infinit workflow) (Gemini realesed a new SYS for reconnections, but I should keep using my own design, because gemini dont give good documentation.)

---

📂 Repository structure

📦 V1.1.9
├── 📁 Notes              # Documentación y referencias
│   ├── 🖼️ workflow.png   # Diagrama del flujo de trabajo
│   ├── 🖼️ workflow22.png
│   ├── 🖼️ workflow21.png
├── 📁 data              
│   ├── 📜 config.json   
│   ├── 📜 current_mode.txt
│   ├── 📜 history_tool.txt
│   ├── 📜 status.txt
│   ├── 📜 voice.txt
│   └── 📂 instructions
│       ├── 📜 lentes_ai.txt
│       ├── 📜 conversacional.txt
│       └── 📜 agente_psicologo.txt
├── 📁 funciones                
│   ├── 📜 audio_def.py        
│   ├── 📜 basic_def.py       
│   ├── 📜 pya_def.py
│   ├── 📜 config.py
│   └── 📜 video_def.py
├── 📜 main.py
├── 📜 exe.py
└── 📜 requirements.txt

---

Workflow's Used and Develop time ✴

System	Problem Description	Workflow Description	Time Used
Re-connection + Memory	💾 Lack of native history context saving.	⚙️ Implemented a function to save and retrieve history context using model instructions.	15 Days, 30 Hours
UI	🖥️ Electron UI incompatible with current backend. TkInter and PyQt were too basic.	🎨 Developed an attractive CMD ASCII UI using Colorama.	5 Days, 7 Hours
Modularization	🧱 2700 Lines of code, that make the coding IMPOSSIBLE.	📂 Finnaly, i decided modularize the code, with an advanced structure of folders and stuff, it was so hard, I litteraly re-make all the code by zero	10 Days, 27 Hours
Language Code	🗣️ It was a hearth hurt that the voice didnt sound good in spanish, and when I finnaly got the API implement, A lot of pydantic errors	🐛 I solve it, by found an error with the class name deprected to discard, soon. I was using it so i can't pass the parameter language_code "" finnaly i solved it	10 Days, 11 Hours

Notes

📸 Papers that I made during develop phase

A continuación, algunas imágenes relacionadas con el proyecto:

---

🚀 How to use

1. Clone the repository:

git clone https://github.com/OptionIA/Gemini-AI-lens-2025-/tree/main

2. Install dependencies:

pip install -r requirements.txt

3. Run the program:

python Code_Source/exe.py # Change folder name, due mayus

4. Go to configuration and update your gemini api keys:

5. Setup your input audio and output audio devices in config

6. Start the mode to use

⚖️ Licencia

🚫 READ LICENSE BEFORE USE – READ LICENSE BEFORE USE

---