Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -108,31 +108,88 @@ def yolov9_inference(model_id, img_path=None, vid_path=None, tracking_algorithm
|
|
108 |
plt.tight_layout() # Ensure the entire plot fits into the figure area
|
109 |
#ax.set_facecolor('#D3D3D3')
|
110 |
elif output_extension.lower() in vid_extensions:
|
111 |
-
output_video = output_path # Load the video file here
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
112 |
output_image = None
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
plt.style.use("ggplot")
|
114 |
fig, ax = plt.subplots(figsize=(10, 6))
|
115 |
-
|
116 |
-
|
117 |
-
sns.lineplot(ax
|
118 |
-
|
119 |
ax.set_title('Number of Objects over Seconds', fontsize=20, pad=20) # Increase padding for the title
|
120 |
ax.set_xlabel('Second', fontsize=16) # Increase font size
|
121 |
ax.set_ylabel('Object Count', fontsize=16) # Increase font size
|
122 |
ax.tick_params(axis='x', labelsize=12) # Increase label size for x-axis
|
123 |
ax.tick_params(axis='y', labelsize=12) # Increase label size for y-axis
|
124 |
-
|
125 |
# Add grid but make it lighter and put it behind bars
|
126 |
ax.grid(True, linestyle=':', linewidth=0.6, color='gray', alpha=0.6)
|
127 |
ax.set_axisbelow(True)
|
128 |
-
|
129 |
# Change the background color to a lighter shade
|
130 |
ax.set_facecolor('#F0F0F0')
|
131 |
-
|
132 |
# Add a legend with a smaller font size
|
133 |
ax.legend(fontsize=10)
|
|
|
|
|
134 |
|
135 |
-
plt.tight_layout() # Ensure the entire
|
136 |
return output_image, output_video, fig
|
137 |
|
138 |
|
|
|
108 |
plt.tight_layout() # Ensure the entire plot fits into the figure area
|
109 |
#ax.set_facecolor('#D3D3D3')
|
110 |
elif output_extension.lower() in vid_extensions:
|
111 |
+
# output_video = output_path # Load the video file here
|
112 |
+
# output_image = None
|
113 |
+
# plt.style.use("ggplot")
|
114 |
+
# fig, ax = plt.subplots(figsize=(10, 6))
|
115 |
+
# #for label in labels:
|
116 |
+
# #df_label = frame_counts_df[frame_counts_df['label'] == label]
|
117 |
+
# sns.lineplot(ax = ax, data = frame_counts_df, x = 'frame', y = 'count', hue = 'label', palette=palette,linewidth=2.5)
|
118 |
+
|
119 |
+
# ax.set_title('Number of Objects over Seconds', fontsize=20, pad=20) # Increase padding for the title
|
120 |
+
# ax.set_xlabel('Second', fontsize=16) # Increase font size
|
121 |
+
# ax.set_ylabel('Object Count', fontsize=16) # Increase font size
|
122 |
+
# ax.tick_params(axis='x', labelsize=12) # Increase label size for x-axis
|
123 |
+
# ax.tick_params(axis='y', labelsize=12) # Increase label size for y-axis
|
124 |
+
|
125 |
+
# # Add grid but make it lighter and put it behind bars
|
126 |
+
# ax.grid(True, linestyle=':', linewidth=0.6, color='gray', alpha=0.6)
|
127 |
+
# ax.set_axisbelow(True)
|
128 |
+
|
129 |
+
# # Change the background color to a lighter shade
|
130 |
+
# ax.set_facecolor('#F0F0F0')
|
131 |
+
|
132 |
+
# # Add a legend with a smaller font size
|
133 |
+
# ax.legend(fontsize=10)
|
134 |
+
|
135 |
+
# plt.tight_layout() # Ensure the entire
|
136 |
+
output_video = output_path
|
137 |
output_image = None
|
138 |
+
|
139 |
+
# Interpolation preprocessing
|
140 |
+
interpolated_data = []
|
141 |
+
|
142 |
+
labels = frame_counts_df['label'].unique()
|
143 |
+
for label in labels:
|
144 |
+
df_label = frame_counts_df[frame_counts_df['label'] == label]
|
145 |
+
|
146 |
+
# Sort data by frame to ensure smooth interpolation
|
147 |
+
df_label = df_label.sort_values('frame')
|
148 |
+
|
149 |
+
# Original data points
|
150 |
+
x = df_label['frame']
|
151 |
+
y = df_label['count']
|
152 |
+
|
153 |
+
# Check if we have enough points for interpolation
|
154 |
+
if len(x) > 1:
|
155 |
+
# Create spline interpolation
|
156 |
+
x_smooth = np.linspace(x.min(), x.max(), 500)
|
157 |
+
spline = make_interp_spline(x, y, k=3) # Cubic spline interpolation
|
158 |
+
y_smooth = spline(x_smooth)
|
159 |
+
|
160 |
+
# Append the smoothed data to the list
|
161 |
+
interpolated_data.append(pd.DataFrame({'frame': x_smooth, 'count': y_smooth, 'label': label}))
|
162 |
+
|
163 |
+
# Concatenate all interpolated data into a single DataFrame
|
164 |
+
if interpolated_data:
|
165 |
+
interpolated_df = pd.concat(interpolated_data)
|
166 |
+
else:
|
167 |
+
interpolated_df = pd.DataFrame(columns=['frame', 'count', 'label'])
|
168 |
+
|
169 |
plt.style.use("ggplot")
|
170 |
fig, ax = plt.subplots(figsize=(10, 6))
|
171 |
+
|
172 |
+
# Plot using Seaborn
|
173 |
+
sns.lineplot(ax=ax, data=interpolated_df, x='frame', y='count', hue='label', palette=palette, linewidth=2.5)
|
174 |
+
|
175 |
ax.set_title('Number of Objects over Seconds', fontsize=20, pad=20) # Increase padding for the title
|
176 |
ax.set_xlabel('Second', fontsize=16) # Increase font size
|
177 |
ax.set_ylabel('Object Count', fontsize=16) # Increase font size
|
178 |
ax.tick_params(axis='x', labelsize=12) # Increase label size for x-axis
|
179 |
ax.tick_params(axis='y', labelsize=12) # Increase label size for y-axis
|
180 |
+
|
181 |
# Add grid but make it lighter and put it behind bars
|
182 |
ax.grid(True, linestyle=':', linewidth=0.6, color='gray', alpha=0.6)
|
183 |
ax.set_axisbelow(True)
|
184 |
+
|
185 |
# Change the background color to a lighter shade
|
186 |
ax.set_facecolor('#F0F0F0')
|
187 |
+
|
188 |
# Add a legend with a smaller font size
|
189 |
ax.legend(fontsize=10)
|
190 |
+
|
191 |
+
plt.tight_layout() # Ensure the entire plot is visible
|
192 |
|
|
|
193 |
return output_image, output_video, fig
|
194 |
|
195 |
|