EmoBot v2 — Emotion-Aware Multilingual Chatbot

An emotionally intelligent chatbot that detects the user's emotion from their message using XLM-RoBERTa and generates empathetic, tone-aligned responses via Google Gemini.

---

Features

• Real-time emotion detection — XLM-RoBERTa fine-tuned on GoEmotions (28 emotion labels)
• Empathetic AI responses — Gemini adapts tone based on detected emotion (sadness → warm, anger → calm, joy → upbeat, ...)
• Multilingual — Detects emotion and responds in the user's language
• Voice input & output — Speech-to-text (Web Speech API + Whisper fallback) and text-to-speech (Edge TTS + browser fallback)
• Automatic model fallback — If Gemini 2.5 Flash quota is exhausted, falls back to Gemma 3 4B seamlessly
• Emotion badges — Detected emotion displayed on both user and bot messages
• Chat actions — Copy, like/dislike feedback, download conversation as JSON, new chat
• Dark mode — Toggle between light and dark themes
• Glassmorphism UI — Frosted glass cards, animated floating orbs, gradient accents
• Security hardened — CSP headers, rate limiting, input sanitization

---

How It Works

User Message
    │
    ▼
┌──────────────────────┐
│  XLM-RoBERTa         │──→ Detected emotion (e.g. "joy", "sadness", "anger")
│  Emotion Classifier  │    + confidence scores (top 3)
└──────────────────────┘
    │
    ▼
┌──────────────────────┐
│  Google Gemini        │──→ Emotionally aligned response
│  (2.5 Flash / Gemma) │    in the user's language
└──────────────────────┘
    │
    ▼
Chat UI displays response + emotion badge

---

Tech Stack

Layer	Technology
Backend	Python, Flask
Emotion Model	XLM-RoBERTa (GoEmotions, 28 labels)
Response Gen	Gemini 2.5 Flash (fallback: Gemma 3 4B)
TTS	Edge TTS (14 language voices)
STT	Web Speech API + OpenAI Whisper
Frontend	HTML, CSS, JavaScript
Database	MongoDB (optional, in-memory fallback)

---

Project Structure

EmoBot/
├── app.py                        # Flask backend (API, emotion detection, Gemini, TTS/STT)
├── requirements.txt              # Python dependencies
├── .env.example                  # Environment variable template
├── .gitignore
├── README.md
├── xlm-roberta.ipynb             # Model training notebook
│
├── static/
│   ├── css/
│   │   └── main.css              # Full stylesheet (glassmorphism, animations, dark mode)
│   ├── js/
│   │   ├── chat.js               # Chat logic (send, receive, TTS, STT, feedback, download)
│   │   └── home.js               # Landing page (emotion pills, scroll reveal, dark mode)
│   └── models/
│       └── xlm-roberta_emotion_model/
│           ├── model.safetensors  # Fine-tuned weights (~1.1 GB, Git LFS)
│           ├── config.json
│           ├── tokenizer_config.json
│           ├── sentencepiece.bpe.model
│           └── special_tokens_map.json
│
└── templates/
    ├── index.html                # Landing page
    ├── chat.html                 # Chat interface
    ├── 404.html                  # Not found page
    └── 500.html                  # Server error page

---

Setup

1. Clone & pull model weights

git clone https://github.com/anVSS1/emobot.git
cd emobot
git lfs install
git lfs pull

2. Create virtual environment

python -m venv .venv
# Windows
.venv\Scripts\activate
# macOS / Linux
source .venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment

cp .env.example .env

Edit .env and add your Google AI Studio API key:

GEMINI_API_KEY=your_key_here

5. Run

python -u app.py

Open http://localhost:5000 and start chatting.

---

Environment Variables

Variable	Default	Description
GEMINI_API_KEY	(required)	Google AI Studio API key
GEMINI_MODEL	gemini-2.5-flash	Primary generation model
FALLBACK_MODEL	gemma-3-4b-it	Fallback model when quota exhausted
SECRET_KEY	(auto-generated)	Flask session secret
MONGODB_URI	mongodb://localhost:27017/	MongoDB connection string
HOST	127.0.0.1	Server bind address
PORT	5000	Server port
FLASK_DEBUG	false	Enable Flask debug mode

---

Supported Emotions (28)

admiration · amusement · anger · annoyance · approval · caring · confusion · curiosity · desire · disappointment · disapproval · disgust · embarrassment · excitement · fear · gratitude · grief · joy · love · nervousness · optimism · pride · realization · relief · remorse · sadness · surprise · neutral

---

Team

Name	Links
Anass Ouzaouit	GitHub · LinkedIn · X
Mohamed Kannoun	LinkedIn · X

---

License

MIT