Spaces:
Running
Running
File size: 10,035 Bytes
3bb804c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 |
"""
PC Scanner Node - Automatically scans through all principal components
Creates a contact sheet showing what each PC controls
"""
import numpy as np
from PyQt6 import QtGui
import cv2
import __main__
BaseNode = __main__.BaseNode
QtGui = __main__.QtGui
class PCScannerNode(BaseNode):
"""
Systematically scans through all PCs to visualize their effects.
Creates a grid showing: [PC0-, PC0+, PC1-, PC1+, ...]
"""
NODE_CATEGORY = "AI / Physics"
NODE_COLOR = QtGui.QColor(220, 180, 100)
def __init__(self, scan_amplitude=2.0, grid_cols=4):
super().__init__()
self.node_title = "PC Scanner"
self.inputs = {
'latent_in': 'spectrum',
'reconstructed_image': 'image', # Feedback from iFFT
'scan_speed': 'signal', # How fast to scan
'trigger': 'signal', # Start scan
'amplitude': 'signal' # How much to modify each PC
}
self.outputs = {
'latent_out': 'spectrum', # Modified latent for current scan
'contact_sheet': 'image', # The full grid
'current_pc': 'signal', # Which PC we're scanning
'progress': 'signal' # 0-1 scan progress
}
self.scan_amplitude = float(scan_amplitude)
self.grid_cols = int(grid_cols)
# Scanning state
self.is_scanning = False
self.scan_index = 0 # Which PC we're currently scanning
self.scan_direction = 1 # 1 for +, -1 for -
self.frame_counter = 0
self.frames_per_scan = 30 # How many frames to wait per PC
# Storage
self.base_latent = None
self.current_latent = None
self.captured_images = {} # {(pc_idx, direction): image}
self.contact_sheet = None
# Dimensions
self.cell_size = 64
def step(self):
# Get inputs
latent_in = self.get_blended_input('latent_in', 'first')
reconstructed = self.get_blended_input('reconstructed_image', 'mean')
trigger = self.get_blended_input('trigger', 'sum') or 0.0
scan_speed = self.get_blended_input('scan_speed', 'sum')
amplitude_signal = self.get_blended_input('amplitude', 'sum')
if amplitude_signal is not None:
amplitude = amplitude_signal * 5.0
else:
amplitude = self.scan_amplitude
if scan_speed is not None:
self.frames_per_scan = max(5, int(30 / (scan_speed + 0.1)))
# Store base latent
if latent_in is not None and self.base_latent is None:
self.base_latent = latent_in.copy()
self.current_latent = latent_in.copy()
# Trigger scan
if trigger > 0.5 and not self.is_scanning:
self.start_scan()
# Scanning logic
if self.is_scanning and self.base_latent is not None:
self.frame_counter += 1
# Capture reconstructed image
if reconstructed is not None and self.frame_counter > 5: # Wait a few frames for stabilization
key = (self.scan_index, self.scan_direction)
if key not in self.captured_images:
# Resize and store
img_resized = cv2.resize(reconstructed, (self.cell_size, self.cell_size))
self.captured_images[key] = img_resized.copy()
# Time to move to next scan?
if self.frame_counter >= self.frames_per_scan:
self.advance_scan()
# Generate current modified latent
self.current_latent = self.base_latent.copy()
if self.scan_index < len(self.base_latent):
self.current_latent[self.scan_index] += amplitude * self.scan_direction
# Build contact sheet
if len(self.captured_images) > 0:
self.build_contact_sheet()
def start_scan(self):
"""Start a new scan"""
self.is_scanning = True
self.scan_index = 0
self.scan_direction = -1 # Start with negative
self.frame_counter = 0
self.captured_images = {}
print("PC Scanner: Starting scan...")
def advance_scan(self):
"""Move to next PC/direction"""
self.frame_counter = 0
if self.scan_direction == -1:
# Switch to positive
self.scan_direction = 1
else:
# Move to next PC
self.scan_direction = -1
self.scan_index += 1
# Check if scan complete
if self.scan_index >= len(self.base_latent):
self.is_scanning = False
self.current_latent = self.base_latent.copy()
print(f"PC Scanner: Scan complete! Captured {len(self.captured_images)} images.")
def build_contact_sheet(self):
"""Build the grid visualization"""
num_pcs = len(self.base_latent) if self.base_latent is not None else 8
# Calculate grid dimensions
# Each PC gets 2 cells (- and +)
total_cells = num_pcs * 2
rows = (total_cells + self.grid_cols - 1) // self.grid_cols
# Create canvas
canvas_width = self.grid_cols * self.cell_size
canvas_height = rows * self.cell_size
canvas = np.zeros((canvas_height, canvas_width, 3), dtype=np.float32)
# Fill grid
cell_idx = 0
for pc_idx in range(num_pcs):
for direction in [-1, 1]:
key = (pc_idx, direction)
row = cell_idx // self.grid_cols
col = cell_idx % self.grid_cols
y_start = row * self.cell_size
x_start = col * self.cell_size
if key in self.captured_images:
img = self.captured_images[key]
# Ensure correct shape
if img.ndim == 2:
img = np.stack([img, img, img], axis=-1)
elif img.shape[2] == 1:
img = np.repeat(img, 3, axis=2)
canvas[y_start:y_start+self.cell_size,
x_start:x_start+self.cell_size] = img
else:
# Empty cell - draw placeholder
cv2.rectangle(canvas, (x_start, y_start),
(x_start+self.cell_size-1, y_start+self.cell_size-1),
(0.2, 0.2, 0.2), 1)
# Label
label = f"PC{pc_idx}" + ("-" if direction == -1 else "+")
cv2.putText(canvas, label, (x_start+2, y_start+12),
cv2.FONT_HERSHEY_SIMPLEX, 0.3, (1, 1, 1), 1)
# Highlight current scan position
if self.is_scanning and pc_idx == self.scan_index and direction == self.scan_direction:
cv2.rectangle(canvas, (x_start, y_start),
(x_start+self.cell_size-1, y_start+self.cell_size-1),
(0, 1, 0), 2)
cell_idx += 1
self.contact_sheet = canvas
def get_output(self, port_name):
if port_name == 'latent_out':
return self.current_latent
elif port_name == 'contact_sheet':
return self.contact_sheet
elif port_name == 'current_pc':
return float(self.scan_index) if self.is_scanning else -1.0
elif port_name == 'progress':
if self.base_latent is not None and self.is_scanning:
total = len(self.base_latent) * 2
current = self.scan_index * 2 + (0 if self.scan_direction == -1 else 1)
return current / total
return 0.0
return None
def get_display_image(self):
if self.contact_sheet is not None:
# Display the contact sheet
img = (np.clip(self.contact_sheet, 0, 1) * 255).astype(np.uint8)
h, w = img.shape[:2]
return QtGui.QImage(img.data, w, h, w*3, QtGui.QImage.Format.Format_RGB888)
else:
# Show status
img = np.zeros((256, 256, 3), dtype=np.uint8)
if self.is_scanning:
status = f"Scanning PC{self.scan_index}{'-' if self.scan_direction == -1 else '+'}"
progress = int(self.get_output('progress') * 100)
cv2.putText(img, status, (10, 128), cv2.FONT_HERSHEY_SIMPLEX,
0.6, (0,255,0), 2)
cv2.putText(img, f"{progress}%", (10, 160), cv2.FONT_HERSHEY_SIMPLEX,
0.5, (255,255,255), 1)
# Progress bar
bar_width = int(256 * self.get_output('progress'))
cv2.rectangle(img, (0, 240), (bar_width, 256), (0,255,0), -1)
else:
cv2.putText(img, "Ready to scan", (10, 128), cv2.FONT_HERSHEY_SIMPLEX,
0.6, (255,255,255), 1)
cv2.putText(img, "Send trigger signal", (10, 160), cv2.FONT_HERSHEY_SIMPLEX,
0.4, (200,200,200), 1)
return QtGui.QImage(img.data, 256, 256, 256*3, QtGui.QImage.Format.Format_RGB888)
def get_config_options(self):
return [
("Scan Amplitude", "scan_amplitude", self.scan_amplitude, None),
("Grid Columns", "grid_cols", self.grid_cols, None)
] |