Spaces:
Runtime error
Runtime error
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from transformers import pipeline
|
3 |
+
import cv2
|
4 |
+
from PIL import Image
|
5 |
+
import io
|
6 |
+
import scipy
|
7 |
+
import torch
|
8 |
+
import time
|
9 |
+
|
10 |
+
|
11 |
+
|
12 |
+
def video_to_descriptions(video, target_language="en"):
|
13 |
+
|
14 |
+
start_time = time.time()
|
15 |
+
print("START TIME = ", start_time)
|
16 |
+
|
17 |
+
ImgToText = pipeline("image-to-text", model="Salesforce/blip-image-captioning-large")
|
18 |
+
Summarize = pipeline("summarization", model="tuner007/pegasus_summarizer")
|
19 |
+
translator = pipeline("translation", model=f"Helsinki-NLP/opus-mt-en-{target_language}")
|
20 |
+
audio = pipeline("text-to-speech", model="suno/bark-small")
|
21 |
+
|
22 |
+
voice_preset = f"v2/{target_language}_speaker_1"
|
23 |
+
|
24 |
+
cap = cv2.VideoCapture(video)
|
25 |
+
fps = int(cap.get(cv2.CAP_PROP_FPS))
|
26 |
+
|
27 |
+
descriptions = []
|
28 |
+
frame_count = 0
|
29 |
+
|
30 |
+
while True:
|
31 |
+
ret, frame = cap.read()
|
32 |
+
if not ret:
|
33 |
+
break
|
34 |
+
|
35 |
+
|
36 |
+
if frame_count % (fps * 2) == 0:
|
37 |
+
|
38 |
+
frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
39 |
+
|
40 |
+
pil_img = Image.fromarray(frame_rgb)
|
41 |
+
|
42 |
+
outputs = ImgToText(pil_img)
|
43 |
+
description = outputs[0]['generated_text']
|
44 |
+
descriptions.append(description)
|
45 |
+
print(str(frame_count) + " : " + outputs[0]['generated_text'])
|
46 |
+
|
47 |
+
frame_count += 1
|
48 |
+
|
49 |
+
cap.release()
|
50 |
+
|
51 |
+
concatenated_description = " ".join(descriptions)
|
52 |
+
summarized_description = Summarize(concatenated_description, max_length=31)[0]["summary_text"]
|
53 |
+
print("SUMMARIZATION : " + summarized_description)
|
54 |
+
|
55 |
+
translated_text = translator(summarized_description)[0]["translation_text"]
|
56 |
+
print("TRANSLATION : " + translated_text)
|
57 |
+
|
58 |
+
audio_file = audio(translated_text)
|
59 |
+
|
60 |
+
output_path = "./bark_out.wav"
|
61 |
+
scipy.io.wavfile.write(output_path, data=audio_file["audio"][0], rate=audio_file["sampling_rate"])
|
62 |
+
|
63 |
+
stop_time = time.time()
|
64 |
+
|
65 |
+
print("EXECUTION TIME = ", stop_time - start_time)
|
66 |
+
return output_path
|
67 |
+
|
68 |
+
language_dropdown = gr.Dropdown(
|
69 |
+
["en", "fr", "de", "es"], label="[MANDATORY] Language", info="The Voice's Language"
|
70 |
+
)
|
71 |
+
|
72 |
+
iface = gr.Interface(
|
73 |
+
fn=video_to_descriptions,
|
74 |
+
inputs=[gr.Video(label="Video to Upload", info="The Video"), language_dropdown],
|
75 |
+
outputs="audio",
|
76 |
+
live=False
|
77 |
+
)
|
78 |
+
|
79 |
+
if __name__ == "__main__":
|
80 |
+
iface.launch()
|