bpiyush commited on
Commit
84703e0
·
1 Parent(s): f21b410

Adds app.py for AVTime dataset

Browse files
Files changed (1) hide show
  1. app.py +75 -0
app.py ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Streamlit demo to visualize AVTime Dataset."""
2
+ import os
3
+ from os.path import join, exists, dirname, abspath, basename
4
+
5
+ import numpy as np
6
+ import pandas as pd
7
+ import streamlit as st
8
+
9
+ import warnings
10
+ warnings.simplefilter(action='ignore')
11
+
12
+ curr_filepath = abspath(__file__)
13
+ repo_path = dirname(dirname(curr_filepath))
14
+
15
+
16
+ def make_grid(cols,rows):
17
+ grid = [0]*cols
18
+ for i in range(cols):
19
+ with st.container():
20
+ grid[i] = st.columns(rows)
21
+ return grid
22
+
23
+
24
+ def read_spreadsheet(sheet_id, gid, **kwargs):
25
+ BASE_URL = 'https://docs.google.com/spreadsheets/d/'
26
+ df = pd.read_csv(BASE_URL + sheet_id + f'/export?gid={gid}&format=csv', **kwargs)
27
+ # drop all rows which have atleast 1 NaN value
28
+ df = df.dropna(axis=0)
29
+ return df
30
+
31
+
32
+ SHEET_ID = "1ejT6SWbihrYzfB_npb0WS2UHd8TIxPZkMpqbJ-M_hfw"
33
+ GID = "0"
34
+
35
+
36
+ if __name__ == "__main__":
37
+ # Add a title
38
+ st.set_page_config(layout="wide")
39
+ st.title("Clips from AVTime Dataset 🎬")
40
+
41
+ if "df" not in st.session_state:
42
+ # Read spreadsheet
43
+ df = read_spreadsheet(sheet_id=SHEET_ID, gid=GID, on_bad_lines='skip')
44
+ st.session_state.df = df
45
+ else:
46
+ df = st.session_state.df
47
+
48
+ st.markdown(f"**Total number of relevant clips**: {len(df)}", unsafe_allow_html=True)
49
+
50
+ reload_button = st.button("Reload")
51
+ NUM = 9
52
+ indices = np.random.randint(0, len(st.session_state.df), NUM)
53
+ if reload_button:
54
+ indices = np.random.randint(0, len(st.session_state.df), NUM)
55
+
56
+ # Create a grid of videos
57
+ grid = make_grid(3, 3)
58
+ per_video_width = 360
59
+ per_video_height = 240
60
+
61
+ for idx, index in enumerate(indices):
62
+ sample = st.session_state.df.iloc[index].to_dict()
63
+ video_id = sample["video_id"]
64
+ stime, etime = sample["start_time"], sample["end_time"]
65
+ duration = etime - stime
66
+ desc = sample["description"]
67
+
68
+ i, j = idx // 3, idx % 3
69
+ grid[i][j].caption(f"Segment duration: {duration}")
70
+ url = f"https://www.youtube.com/embed/{video_id}?start={int(stime)}&end={int(etime)}"
71
+ html_code = f"""
72
+ <iframe height="{per_video_height}" width="{per_video_width}" src="{url}" frameborder="0" allowfullscreen></iframe>
73
+ """
74
+ grid[i][j].markdown(html_code, unsafe_allow_html=True)
75
+ grid[i][j].caption(f"{desc}")