File size: 759 Bytes
caae3e0
3afe20e
caae3e0
 
 
 
2337a39
 
 
 
 
 
 
 
 
 
0bb8b0b
cc1887c
ff3f385
cc1887c
283fe0a
cc1887c
 
caae3e0
0bb8b0b
 
caae3e0
0bb8b0b
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
import gradio as gr
import numpy as np
from huggingface_hub import InferenceClient

client = InferenceClient("models/microsoft/trocr-base-handwritten")

def sepia(input_img):
    sepia_filter = np.array([
        [0.393, 0.769, 0.189],
        [0.349, 0.686, 0.168],
        [0.272, 0.534, 0.131]
    ])
    sepia_img = input_img.dot(sepia_filter.T)
    sepia_img /= sepia_img.max()
    return sepia_img

sepia_interface = gr.Interface(sepia, gr.Image(), "image")

## https://www.gradio.app/docs/gradio/blocks
with gr.Blocks() as generated_output:
    inp = sepia_interface
    out = gr.Textbox()


demoApp = gr.TabbedInterface([sepia_interface, generated_output], ["RGB Sepia Filter", "Handwritten to Text"])

if __name__ == "__main__":
    demoApp.launch()