---
title: AI Child Behavior Assessment
emoji: 🧒
colorFrom: blue
colorTo: green
sdk: streamlit
app_file: app.py
pinned: false
sdk_version: 1.42.0
---

AI Child Behavior Assessment

Multimodal AI-powered tool for analyzing child emotions and speech patterns

🚀 Live Demo: Hugging Face Spaces

📌 Overview

The AI Child Behavior Assessment app is designed to analyze children’s emotional and speech patterns using multimodal AI models. It integrates:
	•	Facial Emotion Recognition (DeepFace) 🧑‍🎨
	•	Speech Analysis & Transcription (Wav2Vec2) 🎙️
	•	Multimodal Analysis (Video + Audio combined) 🎥 + 🔊

This tool helps in early mental health screening and behavioral assessments for research, caregivers, and psychologists.

✨ Features

✅ 1. Video-Based Emotion Analysis
	•	Uses DeepFace AI to detect facial expressions and emotions.
	•	Processes video frames to determine dominant emotions.
	•	Generates a visual summary of detected emotions.

✅ 2. Audio-Based Speech & Tone Analysis
	•	Uses Wav2Vec2 to transcribe spoken words.
	•	Applies speech emotion recognition to assess tone and sentiment.
	•	Includes noise reduction for clearer transcriptions.

✅ 3. Multimodal Analysis (Video + Audio Combined)
	•	Extracts both visual and speech cues to detect behavior patterns.
	•	Compares facial emotions with speech tone to identify inconsistencies.
	•	Provides comprehensive insights into child behavior.

✅ 4. Data Visualization
	•	Displays emotion distribution over time using bar charts 📊.
	•	Generates speech vs. video emotion comparison charts 🏆.

🔧 How to Use

1️⃣ Select an Analysis Mode:
	•	Upload a video file for emotion recognition.
	•	Upload an audio file for speech analysis.
	•	Upload a video + audio file for multimodal analysis.

2️⃣ Click “Analyze” to process the uploaded file.
3️⃣ View Results:
	•	Detected emotions, speech transcription, and analysis insights will be displayed.

📂 Supported File Formats

Analysis Type	Supported Formats
Video 🎥	MP4, AVI, MOV
Audio 🎙️	WAV, MP3
Multimodal (Video + Audio)	MP4, MOV

🔍 Future Improvements

🚀 Planned Enhancements:
✅ Real-time emotion tracking for live video
✅ AI-driven predictive analysis for behavioral trends
✅ Integration with clinical psychology datasets for validation
✅ More advanced multimodal deep learning models

📜 Citation & Acknowledgment

If you use this tool in research or projects, please cite:
Durganihantri Low – AI Child Behavior Assessment (2025)
🌐 Hugging Face Spaces

👨‍💻 Contact & Contributions

Have suggestions or want to contribute? Contact me:
📧 Email: ask@durganihantri.com
🔗 LinkedIn: http://linkedin.com/in/durganihantri