🛡️ PPE Detection with YOLO

Upload a video to detect Personal Protective Equipment using your custom YOLO model
0.1 1

📋 Model Information

Available classes: Glass, Gloves, Goggles, Helmet, No-Helmet, No-Vest, Person, Safety-Boot, Safety-Vest, Vest, helmet, no helmet, no vest, no_helmet, no_vest, protective_suit, vest, worker

🔗 API Usage

cURL Example (2-step process):

# Step 1: Upload video file
VIDEO_HANDLE=$(curl -X POST \
  -F "files=@your_video.mp4" \
  https://rohannair-ppe.hf.space/upload -s | python3 -c "import sys, json; print(json.load(sys.stdin)[0])")

# Step 2: Make prediction request and poll for results
curl -X POST https://rohannair-ppe.hf.space/gradio_api/call/predict \
  -s -H "Content-Type: application/json" \
  -d "{\"data\": [{\"video\": \"$VIDEO_HANDLE\"}, 0.8]}" \
  | python3 -c "import sys, json; r=json.load(sys.stdin); print(r.get('event_id', ''))" \
  | read EVENT_ID; curl -N https://rohannair-ppe.hf.space/gradio_api/call/predict/$EVENT_ID

Python Example (using requests):

import requests
import time

# Step 1: Upload video
with open("video.mp4", "rb") as f:
    upload_response = requests.post(
        "https://rohannair-ppe.hf.space/upload",
        files={"files": ("video.mp4", f, "video/mp4")}
    )
    video_handle = upload_response.json()[0]

# Step 2: Make prediction
response = requests.post(
    "https://rohannair-ppe.hf.space/gradio_api/call/predict",
    json={"data": [{"video": video_handle}, 0.8]},
    headers={"Content-Type": "application/json"}
)

event_id = response.json().get("event_id")

# Step 3: Poll for results
while True:
    result_response = requests.get(
        f"https://rohannair-ppe.hf.space/gradio_api/call/predict/{event_id}"
    )
    result = result_response.json()
    if result.get("status") == "complete":
        print(result["data"])
        break
    time.sleep(2)

Python Example (using Gradio Client - easiest):

from gradio_client import Client

client = Client("https://rohannair-ppe.hf.space/")
result = client.predict(
    "path/to/video.mp4",  # video input
    0.8,  # conf_threshold
    api_name="/predict"
)
print(result)