README.md

metadata

title: PJK GU108
emoji: 🚀
colorFrom: red
colorTo: red
sdk: docker
app_port: 8501
tags:
  - streamlit
pinned: false
short_description: Capstone
license: mit

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

🏫 EduReflect: Analisis Emosi Kelas

EduReflect adalah platform berbasis AI (Computer Vision) yang dirancang untuk menganalisis metrik emosi (afeksi) siswa selama pembelajaran berlangsung melalui rekaman video kelas. Sistem ini mendeteksi wajah siswa secara dinamis dan mengklasifikasikan ekspresi mereka, kemudian memberikan metrik visual serta rekomendasi pedagogis berbasis Pekrun's Control-Value Theory of Achievement Emotions.

🚀 Fitur Utama

• Pelacakan Wajah Dinamis: Menggunakan arsitektur YOLO untuk mendeteksi dan melacak pergerakan wajah secara akurat.
• Klasifikasi Ekspresi: Mengategorikan emosi ke dalam 7 kelas dasar (Happy, Neutral, Surprise, Sad, Anger, Fear, Disgust).
• Dashboard Analitik: Visualisasi interaktif menggunakan Plotly untuk melihat persentase distribusi dan tren timeline emosi kelas per detik.
• Rekomendasi: Memberikan insight dan saran tindakan pedagogis secara otomatis bagi pengajar.

---

⚖️ Attribution & Acknowledgements

🤖 Model Attribution

This Space uses the model yolov11n-face for human face detection and tracking.

• Original Author: deepghs
• Original Model: https://huggingface.co/deepghs/yolo-face/tree/main/yolov11n-face
• License: Model Redistribution Disclaimer License (RDL)

All intellectual property rights and related rights to the Work remain exclusively with the original copyright holder(s). THE WORK IS PROVIDED "AS IS", WITHOUT WARRANTIES OF ANY KIND. The Maintainer explicitly disclaims liability for accuracy, completeness, or fitness for any purpose. All rights remain with the original copyright holder.

📊 Dataset Attribution

The custom emotion classification model used in this repository was trained using a fusion of the following datasets:

1. RAF-DB Dataset

• Dataset: Real-world Affective Faces Database (RAF-DB)
• Author: shuvoalok (Kaggle Upload) / PRIS, Beijing University of Posts and Telecommunications
• Source: https://www.kaggle.com/datasets/shuvoalok/raf-db-dataset
• License/Terms: The RAF database is available for non-commercial research purposes only. All images are obtained from the Internet which are not the property of PRIS. Users agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the images and any portion of derived data.

2. FER2013Plus

• Dataset: Fer2013Plus (Crowd-Sourced Label Distribution by Microsoft)
• Author: subhaditya (Kaggle Porting Code) / Original Microsoft Team (Emad Barsoum, Cha Zhang, Cristian Canton Ferrer, Zhengyou Zhang)
• Source: https://www.kaggle.com/datasets/subhaditya/fer2013plus
• Original Repository: https://github.com/microsoft/FERPlus
• Citation: > Barsoum, E., Zhang, C., Canton Ferrer, C., & Zhang, Z. (2016). Training Deep Networks for Facial Expression Recognition with Crowd-Sourced Label Distribution. ACM International Conference on Multimodal Interaction (ICMI).

---

EduReflect was built as an educational capstone project.