reab5555 commited on
Commit
5272369
·
verified ·
1 Parent(s): 461f83c

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +9 -9
app.py CHANGED
@@ -24,8 +24,8 @@ import tensorflow as tf
24
  print(torch.__version__)
25
  print(torch.version.cuda)
26
 
27
- matplotlib.rcParams['figure.dpi'] = 400
28
- matplotlib.rcParams['savefig.dpi'] = 400
29
 
30
  # Initialize models and other global variables
31
  device = 'cuda'
@@ -356,7 +356,7 @@ def determine_anomalies(mse_values, threshold=5):
356
  return anomalies
357
 
358
 
359
- def plot_mse(df, mse_values, title, color='blue', time_threshold=1, hide_first_n=5):
360
  plt.figure(figsize=(16, 8), dpi=300)
361
  fig, ax = plt.subplots(figsize=(16, 8))
362
 
@@ -514,9 +514,9 @@ def process_video(video_path, desired_fps, batch_size, progress=gr.Progress()):
514
  X, feature_columns, raw_embedding_columns, batch_size=batch_size)
515
 
516
  progress(0.95, "Generating plots")
517
- mse_plot_all = plot_mse(df, mse_all, "Facial Features + Emotions", color='blue', hide_first_n=5)
518
- mse_plot_comp = plot_mse(df, mse_comp, "Facial Features", color='deepskyblue', hide_first_n=5)
519
- mse_plot_raw = plot_mse(df, mse_raw, "Facial Embeddings", color='steelblue', hide_first_n=5)
520
 
521
  emotion_plots = [
522
  plot_mse(df, embedding_anomaly_detection(df[emotion].values.reshape(-1, 1)),
@@ -548,8 +548,8 @@ def process_video(video_path, desired_fps, batch_size, progress=gr.Progress()):
548
 
549
  # Define gallery outputs
550
  gallery_outputs = [
551
- gr.Gallery(label="Most Frequent Person Random Samples", columns=5, rows=2, height="auto"),
552
- gr.Gallery(label="Other Persons Random Samples", columns=5, rows=1, height="auto")
553
  ]
554
 
555
  # Update the Gradio interface
@@ -558,7 +558,7 @@ iface = gr.Interface(
558
  inputs=[
559
  gr.Video(),
560
  gr.Slider(minimum=1, maximum=20, step=1, value=10, label="Desired FPS"),
561
- gr.Slider(minimum=1, maximum=32, step=1, value=8, label="Batch Size")
562
  ],
563
  outputs=[
564
  gr.Textbox(label="Anomaly Detection Results"),
 
24
  print(torch.__version__)
25
  print(torch.version.cuda)
26
 
27
+ matplotlib.rcParams['figure.dpi'] = 500
28
+ matplotlib.rcParams['savefig.dpi'] = 500
29
 
30
  # Initialize models and other global variables
31
  device = 'cuda'
 
356
  return anomalies
357
 
358
 
359
+ def plot_mse(df, mse_values, title, color='blue', time_threshold=1, hide_first_n=2):
360
  plt.figure(figsize=(16, 8), dpi=300)
361
  fig, ax = plt.subplots(figsize=(16, 8))
362
 
 
514
  X, feature_columns, raw_embedding_columns, batch_size=batch_size)
515
 
516
  progress(0.95, "Generating plots")
517
+ mse_plot_all = plot_mse(df, mse_all, "Facial Features + Emotions", color='blue', hide_first_n=2)
518
+ mse_plot_comp = plot_mse(df, mse_comp, "Facial Features", color='deepskyblue', hide_first_n=2)
519
+ mse_plot_raw = plot_mse(df, mse_raw, "Facial Embeddings", color='steelblue', hide_first_n=2)
520
 
521
  emotion_plots = [
522
  plot_mse(df, embedding_anomaly_detection(df[emotion].values.reshape(-1, 1)),
 
548
 
549
  # Define gallery outputs
550
  gallery_outputs = [
551
+ gr.Gallery(label="Most Frequent Person Random Samples", columns=10, rows=2, height="auto"),
552
+ gr.Gallery(label="Other Persons Random Samples", columns=10, rows=1, height="auto")
553
  ]
554
 
555
  # Update the Gradio interface
 
558
  inputs=[
559
  gr.Video(),
560
  gr.Slider(minimum=1, maximum=20, step=1, value=10, label="Desired FPS"),
561
+ gr.Slider(minimum=1, maximum=32, step=1, value=10, label="Batch Size")
562
  ],
563
  outputs=[
564
  gr.Textbox(label="Anomaly Detection Results"),