Spaces:
Running
Running
File size: 12,308 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 |
"""
PixelEaterFliesNode (Artificial Life)
--------------------------------
This node simulates a swarm of "flies" (agents) that
live on, consume, and are guided by an input image.
It is an attempt at "artificial life" [cite: "attempt at artificial life"].
- The "World" is the `image_in` (e.g., from a Mandelbrot node).
- The "Flies" are agents with "dumb" [cite: "dumbflies.py"] logic.
- "Dopamine" [cite: "that.. is driven by dopamine addiction"] is Health.
- "Logic" [cite: "thin sheet of logic"] is: "find and eat the
brightest pixels" [cite: "they can eat the pixels... based on brightness"].
- "Graphics" are inspired by dumbflies.py [cite: "dumbflies.py"].
"""
import numpy as np
import cv2
import time
# --- Magic import block ---
import __main__
BaseNode = __main__.BaseNode
QtGui = __main__.QtGui
# --------------------------
try:
from scipy.ndimage import gaussian_filter
SCIPY_AVAILABLE = True
except ImportError:
SCIPY_AVAILABLE = False
print("Warning: PixelEaterFliesNode requires 'scipy'.")
class FlyAgent:
"""
A single fly. It has a body, logic, and health.
This is the "Dendrite" [cite: "dendrites cause phase space"] or "Scout".
"""
def __init__(self, x, y, grid_size, config):
self.x = float(x)
self.y = float(y)
self.grid_size = grid_size
self.angle = np.random.uniform(0, 2 * np.pi)
self.speed = np.random.uniform(1.0, 3.0)
self.health = 1.0
self.age = np.random.uniform(0, 100) # For wing flapping
# Get config
self.config = config
# --- "Dumb" logic from dumbflies.py ---
self.state = "WANDER"
self.state_timer = 0
self.turn_speed = np.random.uniform(0.1, 0.3)
def _get_pixel_brightness(self, world, x, y):
"""Helper to safely get brightness at a (wrapped) coordinate"""
xi = int(x) % self.grid_size
yi = int(y) % self.grid_size
return world[yi, xi]
def perceive_and_decide(self, world):
"""
The "Thin Logic". The fly "sees" three points in front
of it and steers towards the brightest one (food).
"""
# 1. Perception: Sample 3 points in front
dist = self.config['perception_distance']
center_x = self.x + np.cos(self.angle) * dist
center_y = self.y + np.sin(self.angle) * dist
angle_left = self.angle - np.pi / 4
left_x = self.x + np.cos(angle_left) * dist
left_y = self.y + np.sin(angle_left) * dist
angle_right = self.angle + np.pi / 4
right_x = self.x + np.cos(angle_right) * dist
right_y = self.y + np.sin(angle_right) * dist
food_center = self._get_pixel_brightness(world, center_x, center_y)
food_left = self._get_pixel_brightness(world, left_x, left_y)
food_right = self._get_pixel_brightness(world, right_x, right_y)
# 2. Decision Logic (Steering)
if food_center > food_left and food_center > food_right:
# Food is straight ahead, keep going
self.state = "SEEK"
elif food_left > food_right:
# Food is to the left
self.state = "SEEK"
self.angle -= self.turn_speed
elif food_right > food_left:
# Food is to the right
self.state = "SEEK"
self.angle += self.turn_speed
else:
# No food in sight, wander
self.state = "WANDER"
self.state_timer -= 1
if self.state_timer <= 0:
# Pick a new random turn
self.turn_speed = np.random.uniform(-0.2, 0.2)
self.state_timer = np.random.randint(10, 50)
self.angle += self.turn_speed
self.angle %= (2 * np.pi) # Normalize angle
def update_physics(self):
"""Update position based on angle and speed"""
self.vel_x = np.cos(self.angle) * self.speed
self.vel_y = np.sin(self.angle) * self.speed
self.x += self.vel_x
self.y += self.vel_y
# Wrap around world edges
self.x %= self.grid_size
self.y %= self.grid_size
self.age += 1
def eat_and_live(self, world):
"""
Eat pixels to gain health, decay health over time.
This modifies the "world" (the input image).
"""
xi, yi = int(self.x), int(self.y)
# 1. Eat food (brightness)
food_eaten = world[yi, xi]
if food_eaten > 0.1:
self.health += food_eaten * self.config['food_value']
# Modify the world: "eat" the pixel
world[yi, xi] *= self.config['eat_decay']
# 2. Metabolism
self.health -= self.config['health_decay']
self.health = np.clip(self.health, 0.0, 1.0)
return self.health > 0 # Return True if alive
def draw(self, display_image):
"""
Draw the fly with flapping wings, inspired by dumbflies.py
"""
xi, yi = int(self.x), int(self.y)
# Body
body_color = (int(self.health * 255), 0, 0) # BGR: Red = healthy
cv2.circle(display_image, (xi, yi), 3, body_color, -1)
# Wings
wing_len = 5
wing_angle_base = np.pi / 2 # Perpendicular to body
# Flap wings
wing_flap = np.sin(self.age * 0.5) * 0.5 # Flap angle
# Left Wing
wl_angle = self.angle + wing_angle_base + wing_flap
wl_x = int(xi + np.cos(wl_angle) * wing_len)
wl_y = int(yi + np.sin(wl_angle) * wing_len)
cv2.line(display_image, (xi, yi), (wl_x, wl_y), (200, 200, 200), 1)
# Right Wing
wr_angle = self.angle - wing_angle_base - wing_flap
wr_x = int(xi + np.cos(wr_angle) * wing_len)
wr_y = int(yi + np.sin(wr_angle) * wing_len)
cv2.line(display_image, (xi, yi), (wr_x, wr_y), (200, 200, 200), 1)
class PixelEaterFliesNode(BaseNode):
NODE_CATEGORY = "Artificial Life"
NODE_COLOR = QtGui.QColor(100, 250, 150) # A-Life Green
def __init__(self, num_flies=50, health_decay=0.01, food_value=0.1, perception_distance=10, eat_decay=0.9):
super().__init__()
self.node_title = "Pixel Eater Flies"
self.inputs = {
'image_in': 'image', # The "World"
'reset': 'signal'
}
self.outputs = {
'world_image': 'image', # The "World" + "Flies"
'population': 'signal',
'avg_health': 'signal'
}
if not SCIPY_AVAILABLE:
self.node_title = "Flies (No SciPy!)"
return
self.grid_size = 256 # Default, will adapt to image
# --- Configurable Parameters ---
self.config = {
'num_flies': int(num_flies),
'health_decay': float(health_decay),
'food_value': float(food_value),
'perception_distance': int(perception_distance),
'eat_decay': float(eat_decay) # How much eating darkens a pixel
}
# --- Internal State ---
self.flies = []
self.world = None # This will hold the grayscale, 0-1 "food" map
self.world_vis = None # This will hold the color "drawing"
self.avg_health = 0.0
self._spawn_flies(self.config['num_flies'])
def _spawn_flies(self, count):
for _ in range(count):
x = np.random.randint(0, self.grid_size)
y = np.random.randint(0, self.grid_size)
self.flies.append(FlyAgent(x, y, self.grid_size, self.config))
def _prepare_world(self, img_in):
"""Converts any input image to a 0-1 grayscale float 'food' map"""
# --- Fix for CV_64F crash [cite: "cv2.error... 'depth' is 6 (CV_64F)"] ---
if img_in.dtype != np.float32:
img_in = img_in.astype(np.float32)
if img_in.max() > 1.0:
img_in = img_in / 255.0
# --- End Fix ---
if img_in.shape[0] != self.grid_size or img_in.shape[1] != self.grid_size:
self.grid_size = img_in.shape[0]
# Respawn flies if world size changes
self.flies = []
self._spawn_flies(self.config['num_flies'])
if img_in.ndim == 3:
# Convert to grayscale (Luminance)
world_gray = cv2.cvtColor(img_in, cv2.COLOR_RGB2GRAY)
else:
world_gray = img_in.copy()
return world_gray
def step(self):
if not SCIPY_AVAILABLE:
return
# 1. Get Inputs
reset = self.get_blended_input('reset', 'sum') or 0.0
img_in = self.get_blended_input('image_in', 'first') # Use 'first'
if reset > 0.5:
self.flies = []
self._spawn_flies(self.config['num_flies'])
self.world = None # Force world reload
# 2. Update/Prepare the "World"
if img_in is not None:
# A new world frame is piped in
self.world = self._prepare_world(img_in)
elif self.world is None:
# No world yet, create a default black one
self.world = np.zeros((self.grid_size, self.grid_size), dtype=np.float32)
# Create the color visualization frame
# We draw on a copy so the flies don't "see" their own drawings
self.world_vis = cv2.cvtColor(self.world, cv2.COLOR_GRAY2RGB)
# 3. Update all "Flies"
alive_flies = []
total_health = 0.0
for fly in self.flies:
fly.perceive_and_decide(self.world)
fly.update_physics()
# Eat and check if alive
if fly.eat_and_live(self.world):
fly.draw(self.world_vis) # Draw alive flies
alive_flies.append(fly)
total_health += fly.health
else:
# Fly "died", just don't add it to the list
pass
self.flies = alive_flies
# 4. Handle Reproduction/Respawning
missing_flies = self.config['num_flies'] - len(self.flies)
if missing_flies > 0:
self._spawn_flies(missing_flies)
# 5. Calculate Metrics
if len(self.flies) > 0:
self.avg_health = total_health / len(self.flies)
else:
self.avg_health = 0.0
def get_output(self, port_name):
if port_name == 'world_image':
if self.world_vis is None:
return np.zeros((self.grid_size, self.grid_size, 3), dtype=np.float32)
return self.world_vis
elif port_name == 'population':
return float(len(self.flies))
elif port_name == 'avg_health':
return self.avg_health
return None
def get_display_image(self):
if self.world_vis is None:
img = np.zeros((96, 96, 3), dtype=np.uint8)
else:
img = cv2.resize(self.world_vis, (96, 96), interpolation=cv2.INTER_NEAREST)
# Add Health Bar
health_w = int(self.avg_health * (96 - 4))
cv2.rectangle(img, (2, 96 - 7), (2 + health_w, 96 - 2), (0, 255, 0), -1)
img = np.ascontiguousarray(img)
return QtGui.QImage(img.data, 96, 96, 96*3, QtGui.QImage.Format.Format_BGR888)
def get_config_options(self):
return [
("Num Flies", "num_flies", self.config['num_flies'], None),
("Health Decay", "health_decay", self.config['health_decay'], None),
("Food Value", "food_value", self.config['food_value'], None),
("Perception Distance", "perception_distance", self.config['perception_distance'], None),
("Eat Decay", "eat_decay", self.config['eat_decay'], None),
]
def close(self):
self.flies = []
super().close() |