File size: 1,139 Bytes
7a9d89d 363c085 7a9d89d e323e52 7a9d89d 363c085 53f268c 7a9d89d 53f268c 7a9d89d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
import os
os.system("hub install openpose_body_estimation==1.0.0")
import gradio as gr
import paddlehub as hub
import numpy as np
from PIL import Image
model = hub.Module(name='openpose_body_estimation')
def inference(image):
result = model.predict(image.name)
return Image.fromarray(np.uint8(result['data'])[:,:,::-1]).convert('RGB')
title = "OpenPose"
description = "Gradio demo for OpenPose: Real-time multi-person keypoint detection library for body, face, hands, and foot estimation. To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."
article = "<p style='text-align: center'><a href='https://arxiv.org/abs/1812.08008' target='_blank'>OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields</a> | <a href='https://github.com/CMU-Perceptual-Computing-Lab/openpose' target='_blank'>Github Repo</a></p>"
examples=[['people.jpeg']]
iface = gr.Interface(inference, inputs=gr.inputs.Image(type="file"), outputs=gr.outputs.Image(type="pil"),enable_queue=True,title=title,article=article,description=description,examples=examples)
iface.launch() |