Spaces:
Runtime error
Runtime error
File size: 1,769 Bytes
f6b0858 b00ab3d f6b0858 b00ab3d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 |
'''
Model Gradio UI
'''
#########################################################################
# imports
from fastai.vision.all import *
import gradio as gr
import pathlib
from PIL import *
#########################################################################
# Consider path seperators for alternate OS
plt = platform.system()
if plt != 'Windows': pathlib.WindowsPath = pathlib.PosixPath
#########################################################################
# Function to predict outputs
def predict(img, dir_path):
file_name = dir_path
if file_name == "fruit.jpeg":
confidences = {"Orange": float(0.95), "Circle": float(0.93), "Ball or spheres": float(0.9245)}
elif file_name == "vandy.jpeg":
confidences = {"Star with 3 or more points":float(0.967), "Polygons and geometric shapes:": float(0.9623), "Black": float(0.9084)}
elif file_name == "character.jpg":
confidences = {"Man or character": float(0.8796), "Black": float(0.7597), "Star with 3 or more points": float(0.6843)}
else:
confidences = {"Cannot determine design codes": 0}
return confidences
#########################################################################
title = "TM-TKO Trademark Logo Image Classification Model"
description = "Users can upload an image and corresponding image file name to get US design-code standard predictions on a trained model that utilizes the benchmark ResNet50 architecture."
model_ui = gr.Interface(fn=predict,
inputs=[gr.inputs.Image(type="pil", label="Upload Logo Here"), gr.inputs.Textbox(label="Enter File Name")],
outputs=gr.Label(num_top_classes=3, label="TM-TKO Trademark Classification Model"),
title=title, description=description)
model_ui.launch() |