File size: 12,336 Bytes
338d95d |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 |
"""
CompI Phase 2.A: Audio-to-Image Generation
This module implements multimodal AI art generation that combines:
- Text prompts with style and mood conditioning
- Audio analysis and feature extraction
- Audio-to-text captioning
- Intelligent prompt fusion for enhanced creativity
Features:
- Support for various audio formats (mp3, wav, flac, etc.)
- Real-time audio analysis with tempo, energy, and spectral features
- OpenAI Whisper integration for audio captioning
- Comprehensive metadata logging and filename conventions
- Batch processing capabilities
"""
import os
import sys
import torch
import json
from datetime import datetime
from typing import Dict, List, Optional, Tuple, Union
from pathlib import Path
import logging
# Add project root to path
sys.path.append(os.path.join(os.path.dirname(__file__), '..', '..'))
from diffusers import StableDiffusionPipeline
from PIL import Image
import numpy as np
from src.utils.audio_utils import AudioProcessor, AudioCaptioner, MultimodalPromptFusion, AudioFeatures
from src.utils.logging_utils import setup_logger
from src.utils.file_utils import ensure_directory_exists, generate_filename
# Setup logging
logger = setup_logger(__name__)
class CompIPhase2AAudioToImage:
"""
CompI Phase 2.A: Audio-to-Image Generation System
Combines text prompts with audio analysis to generate contextually rich AI art
"""
def __init__(
self,
model_name: str = "runwayml/stable-diffusion-v1-5",
device: str = "auto",
output_dir: str = "outputs",
whisper_model: str = "base"
):
"""
Initialize the audio-to-image generation system
Args:
model_name: Stable Diffusion model to use
device: Device for inference (auto, cpu, cuda)
output_dir: Directory for saving generated images
whisper_model: Whisper model size for audio captioning
"""
self.model_name = model_name
self.device = self._setup_device(device)
self.output_dir = Path(output_dir)
ensure_directory_exists(self.output_dir)
# Initialize components
self.pipe = None
self.audio_processor = AudioProcessor()
self.audio_captioner = AudioCaptioner(model_size=whisper_model, device=self.device)
self.prompt_fusion = MultimodalPromptFusion()
logger.info(f"Initialized CompI Phase 2.A on {self.device}")
def _setup_device(self, device: str) -> str:
"""Setup and validate device"""
if device == "auto":
device = "cuda" if torch.cuda.is_available() else "cpu"
if device == "cuda" and not torch.cuda.is_available():
logger.warning("CUDA requested but not available, falling back to CPU")
device = "cpu"
return device
def _load_pipeline(self):
"""Lazy load the Stable Diffusion pipeline"""
if self.pipe is None:
logger.info(f"Loading Stable Diffusion model: {self.model_name}")
# Custom safety checker that allows creative content
def dummy_safety_checker(images, **kwargs):
return images, [False] * len(images)
self.pipe = StableDiffusionPipeline.from_pretrained(
self.model_name,
torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
safety_checker=dummy_safety_checker,
requires_safety_checker=False
)
self.pipe = self.pipe.to(self.device)
self.pipe.enable_attention_slicing()
if self.device == "cuda":
self.pipe.enable_model_cpu_offload()
logger.info("Stable Diffusion pipeline loaded successfully")
def analyze_audio(self, audio_path: str, include_caption: bool = True) -> Tuple[AudioFeatures, str]:
"""
Comprehensive audio analysis
Args:
audio_path: Path to audio file
include_caption: Whether to generate audio caption
Returns:
Tuple of (AudioFeatures, audio_caption)
"""
logger.info(f"Analyzing audio: {audio_path}")
# Extract audio features
audio_features = self.audio_processor.analyze_audio_file(audio_path)
# Generate audio caption if requested
audio_caption = ""
if include_caption:
try:
audio_caption = self.audio_captioner.caption_audio(audio_path)
except Exception as e:
logger.warning(f"Audio captioning failed: {e}")
audio_caption = ""
return audio_features, audio_caption
def generate_image(
self,
text_prompt: str,
style: str = "",
mood: str = "",
audio_path: Optional[str] = None,
num_images: int = 1,
height: int = 512,
width: int = 512,
num_inference_steps: int = 30,
guidance_scale: float = 7.5,
seed: Optional[int] = None
) -> List[Dict]:
"""
Generate images with optional audio conditioning
Args:
text_prompt: Base text prompt
style: Art style
mood: Mood/atmosphere
audio_path: Optional path to audio file for conditioning
num_images: Number of images to generate
height: Image height
width: Image width
num_inference_steps: Number of diffusion steps
guidance_scale: Guidance scale for generation
seed: Random seed for reproducibility
Returns:
List of generation results with metadata
"""
self._load_pipeline()
# Analyze audio if provided
audio_features = None
audio_caption = ""
if audio_path and os.path.exists(audio_path):
audio_features, audio_caption = self.analyze_audio(audio_path)
# Create enhanced prompt
if audio_features:
enhanced_prompt = self.prompt_fusion.fuse_prompt_with_audio(
text_prompt, style, mood, audio_features, audio_caption
)
else:
enhanced_prompt = text_prompt
if style:
enhanced_prompt += f", {style}"
if mood:
enhanced_prompt += f", {mood}"
logger.info(f"Generating {num_images} image(s) with prompt: {enhanced_prompt}")
results = []
for i in range(num_images):
# Set up generation parameters
current_seed = seed if seed is not None else torch.seed()
generator = torch.Generator(device=self.device).manual_seed(current_seed)
# Generate image
with torch.autocast(self.device) if self.device == "cuda" else torch.no_grad():
result = self.pipe(
enhanced_prompt,
height=height,
width=width,
num_inference_steps=num_inference_steps,
guidance_scale=guidance_scale,
generator=generator
)
image = result.images[0]
# Create metadata
metadata = {
"timestamp": datetime.now().isoformat(),
"text_prompt": text_prompt,
"style": style,
"mood": mood,
"enhanced_prompt": enhanced_prompt,
"audio_path": audio_path,
"audio_caption": audio_caption,
"generation_params": {
"height": height,
"width": width,
"num_inference_steps": num_inference_steps,
"guidance_scale": guidance_scale,
"seed": current_seed,
"model": self.model_name
},
"device": self.device,
"phase": "2A_audio_to_image"
}
# Add audio features to metadata
if audio_features:
metadata["audio_features"] = audio_features.to_dict()
metadata["audio_tags"] = self.prompt_fusion.generate_audio_tags(audio_features)
# Generate filename
filename = self._generate_filename(
text_prompt, style, mood, current_seed, i + 1,
has_audio=audio_path is not None
)
# Save image and metadata
image_path = self.output_dir / f"{filename}.png"
metadata_path = self.output_dir / f"{filename}_metadata.json"
image.save(image_path)
with open(metadata_path, 'w') as f:
json.dump(metadata, f, indent=2)
results.append({
"image": image,
"image_path": str(image_path),
"metadata_path": str(metadata_path),
"metadata": metadata,
"filename": filename
})
logger.info(f"Generated image {i+1}/{num_images}: {filename}")
return results
def _generate_filename(
self,
prompt: str,
style: str,
mood: str,
seed: int,
variation: int,
has_audio: bool = False
) -> str:
"""Generate descriptive filename following CompI conventions"""
# Create prompt slug (first 5 words)
prompt_words = prompt.lower().replace(',', '').split()[:5]
prompt_slug = "_".join(prompt_words)
# Create style and mood slugs
style_slug = style.replace(" ", "").replace(",", "")[:10] if style else "standard"
mood_slug = mood.replace(" ", "").replace(",", "")[:10] if mood else "neutral"
# Timestamp
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
# Audio indicator
audio_tag = "_AUDIO" if has_audio else ""
# Combine all elements
filename = f"{prompt_slug}_{style_slug}_{mood_slug}_{timestamp}_seed{seed}{audio_tag}_v{variation}"
return filename
def batch_process(
self,
audio_directory: str,
text_prompt: str,
style: str = "",
mood: str = "",
**generation_kwargs
) -> List[Dict]:
"""
Process multiple audio files in batch
Args:
audio_directory: Directory containing audio files
text_prompt: Base text prompt for all generations
style: Art style
mood: Mood/atmosphere
**generation_kwargs: Additional generation parameters
Returns:
List of all generation results
"""
audio_dir = Path(audio_directory)
if not audio_dir.exists():
raise ValueError(f"Audio directory not found: {audio_directory}")
# Find audio files
audio_extensions = {'.mp3', '.wav', '.flac', '.m4a', '.ogg'}
audio_files = [
f for f in audio_dir.iterdir()
if f.suffix.lower() in audio_extensions
]
if not audio_files:
raise ValueError(f"No audio files found in {audio_directory}")
logger.info(f"Processing {len(audio_files)} audio files")
all_results = []
for audio_file in audio_files:
logger.info(f"Processing: {audio_file.name}")
try:
results = self.generate_image(
text_prompt=text_prompt,
style=style,
mood=mood,
audio_path=str(audio_file),
**generation_kwargs
)
all_results.extend(results)
except Exception as e:
logger.error(f"Error processing {audio_file.name}: {e}")
continue
logger.info(f"Batch processing complete: {len(all_results)} images generated")
return all_results
|