mapotofu40/qwen-mlp

DAiSEE Emotion Detection Model

This model detects 4 emotion states (Boredom, Engagement, Confusion, Frustration) from educational videos.

Architecture

• Stage 1: Qwen2.5-VL-7B-Instruct (embedding extraction)
• Stage 2: MLP classifiers (4 separate models)

Usage

Via Inference Endpoint

import requests
import base64

# Encode video
with open("student_video.avi", "rb") as f:
    video_b64 = base64.b64encode(f.read()).decode()

# Send request
response = requests.post(
    "https://YOUR-ENDPOINT.aws.endpoints.huggingface.cloud",
    headers={"Authorization": f"Bearer {YOUR_HF_TOKEN}"},
    json={"inputs": video_b64}
)

predictions = response.json()
print(predictions)