st192011 commited on
Commit
9aa92d2
·
verified ·
1 Parent(s): 6a2ebe8

Update stats_data.py

Browse files
Files changed (1) hide show
  1. stats_data.py +31 -10
stats_data.py CHANGED
@@ -14,21 +14,42 @@ SPEAKER_META = {
14
  }
15
 
16
  def get_indomain_breakdown():
 
17
  data = {
18
  "Speaker": ["M05", "F01", "M01", "M04", "M02", "M03", "F03", "F04"],
19
  "Severity": ["Severe", "Severe", "Moderate", "Moderate", "Mild", "Mild", "Mild", "Mild"],
20
- "Whisper Tiny": ["12.1%", "12.6%", "32.7%", "31.8%", "62.1%", "58.4%", "61.2%", "59.1%"],
21
- "5K Model (Ours)": ["33.1%", "34.2%", "47.2%", "45.6%", "84.5%", "81.8%", "83.5%", "82.8%"],
22
- "10K Model (Ours)": ["25.4%", "24.1%", "44.1%", "41.2%", "79.1%", "77.5%", "79.0%", "78.2%"]
23
  }
24
- return pd.DataFrame(data)
 
 
 
 
 
 
 
 
 
 
 
 
25
 
26
  def get_experimental_summary():
 
27
  data = {
28
- "Experiment Condition": ["In-Domain (Seen Torgo)", "LOSO (Unseen Torgo F01)", "Zero-Shot (UA-Speech F02)"],
29
- "Whisper Tiny (Baseline)": ["41.50%", "12.38%", "4.33%"],
30
- "5K Pure Model Score": ["58.77%", "N/A", "6.19%"],
31
- "10K Triple-Mix Score": ["54.67%", "24.76%", "5.98%"],
32
- "Relative Gain (Best vs Baseline)": ["+41.6%", "+100.0%", "+42.9%"]
33
  }
34
- return pd.DataFrame(data)
 
 
 
 
 
 
 
 
 
 
 
14
  }
15
 
16
  def get_indomain_breakdown():
17
+ # Primary Data for Torgo In-Domain (5K Model is the Accuracy Champion)
18
  data = {
19
  "Speaker": ["M05", "F01", "M01", "M04", "M02", "M03", "F03", "F04"],
20
  "Severity": ["Severe", "Severe", "Moderate", "Moderate", "Mild", "Mild", "Mild", "Mild"],
21
+ "Whisper Tiny": [12.1, 12.6, 32.7, 31.8, 62.1, 58.4, 61.2, 59.1],
22
+ "DSR Lab (5K)": [33.1, 34.2, 47.2, 45.6, 84.5, 81.8, 83.5, 82.8]
 
23
  }
24
+ df = pd.DataFrame(data)
25
+
26
+ # Calculate Gains
27
+ df["Absolute Gain (%)"] = (df["DSR Lab (5K)"] - df["Whisper Tiny"]).round(2)
28
+ df["Relative Improvement (%)"] = (((df["DSR Lab (5K)"] - df["Whisper Tiny"]) / df["Whisper Tiny"]) * 100).round(1)
29
+
30
+ # Format for display
31
+ df["Whisper Tiny"] = df["Whisper Tiny"].astype(str) + "%"
32
+ df["DSR Lab (5K)"] = df["DSR Lab (5K)"].astype(str) + "%"
33
+ df["Absolute Gain (%)"] = "+" + df["Absolute Gain (%)"].astype(str) + "%"
34
+ df["Relative Improvement (%)"] = "+" + df["Relative Improvement (%)"].astype(str) + "%"
35
+
36
+ return df
37
 
38
  def get_experimental_summary():
39
+ # Comparing conditions (In-Domain, LOSO, Zero-Shot)
40
  data = {
41
+ "Condition": ["In-Domain (Seen Torgo)", "LOSO (Unseen Torgo F01)", "Zero-Shot (UA-Speech F02)"],
42
+ "Whisper Tiny": [41.50, 12.38, 4.33],
43
+ "Our Best Score": [58.77, 24.76, 6.19]
 
 
44
  }
45
+ df = pd.DataFrame(data)
46
+
47
+ # Calculate Relative Gain
48
+ df["Relative Gain"] = (((df["Our Best Score"] - df["Whisper Tiny"]) / df["Whisper Tiny"]) * 100).round(1)
49
+
50
+ # Format for display
51
+ df["Whisper Tiny"] = df["Whisper Tiny"].astype(str) + "%"
52
+ df["Our Best Score"] = df["Our Best Score"].astype(str) + "%"
53
+ df["Relative Gain"] = "+" + df["Relative Gain"].astype(str) + "%"
54
+
55
+ return df