VisionScout / clip_model_manager.py
DawnC's picture
Upload 3 files
2fa80e1 verified
raw
history blame
9.1 kB
import torch
import clip
import numpy as np
import logging
import traceback
from typing import List, Dict, Tuple, Optional, Union, Any
from PIL import Image
class CLIPModelManager:
"""
專門管理 CLIP 模型相關的操作,包括模型載入、設備管理、圖像和文本的特徵編碼等核心功能
"""
def __init__(self, model_name: str = "ViT-B/16", device: str = None):
"""
初始化 CLIP 模型管理器
Args:
model_name: CLIP模型名稱,默認為"ViT-B/16"
device: 運行設備,None則自動選擇
"""
self.logger = logging.getLogger(__name__)
self.model_name = model_name
# 設置運行設備
if device is None:
self.device = "cuda" if torch.cuda.is_available() else "cpu"
else:
self.device = device
self.model = None
self.preprocess = None
self._initialize_model()
def _initialize_model(self):
"""
初始化CLIP模型
"""
try:
self.logger.info(f"Initializing CLIP model ({self.model_name}) on {self.device}")
self.model, self.preprocess = clip.load(self.model_name, device=self.device)
self.logger.info("Successfully loaded CLIP model")
except Exception as e:
self.logger.error(f"Error loading CLIP model: {e}")
self.logger.error(traceback.format_exc())
raise
def encode_image(self, image_input: torch.Tensor) -> torch.Tensor:
"""
編碼圖像特徵
Args:
image_input: 預處理後的圖像張量
Returns:
torch.Tensor: 標準化後的圖像特徵
"""
try:
with torch.no_grad():
image_features = self.model.encode_image(image_input)
image_features = image_features / image_features.norm(dim=-1, keepdim=True)
return image_features
except Exception as e:
self.logger.error(f"Error encoding image features: {e}")
self.logger.error(traceback.format_exc())
raise
def encode_text_batch(self, text_prompts: List[str], batch_size: int = 128) -> torch.Tensor:
"""
批量編碼文本特徵,避免CUDA內存問題
Args:
text_prompts: 文本提示列表
batch_size: 批處理大小
Returns:
torch.Tensor: 標準化後的文本特徵
"""
if not text_prompts:
return None
try:
with torch.no_grad():
features_list = []
for i in range(0, len(text_prompts), batch_size):
batch_prompts = text_prompts[i:i+batch_size]
text_tokens = clip.tokenize(batch_prompts).to(self.device)
batch_features = self.model.encode_text(text_tokens)
batch_features = batch_features / batch_features.norm(dim=-1, keepdim=True)
features_list.append(batch_features)
# 連接所有批次
if len(features_list) > 1:
text_features = torch.cat(features_list, dim=0)
else:
text_features = features_list[0]
return text_features
except Exception as e:
self.logger.error(f"Error encoding text features: {e}")
self.logger.error(traceback.format_exc())
raise
def encode_single_text(self, text_prompts: List[str]) -> torch.Tensor:
"""
編碼單個文本批次的特徵
Args:
text_prompts: 文本提示列表
Returns:
torch.Tensor: 標準化後的文本特徵
"""
try:
with torch.no_grad():
text_tokens = clip.tokenize(text_prompts).to(self.device)
text_features = self.model.encode_text(text_tokens)
text_features = text_features / text_features.norm(dim=-1, keepdim=True)
return text_features
except Exception as e:
self.logger.error(f"Error encoding single text batch: {e}")
self.logger.error(traceback.format_exc())
raise
def calculate_similarity(self, image_features: torch.Tensor, text_features: torch.Tensor) -> np.ndarray:
"""
計算圖像和文本特徵之間的相似度
Args:
image_features: 圖像特徵張量
text_features: 文本特徵張量
Returns:
np.ndarray: 相似度分數數組
"""
try:
with torch.no_grad():
similarity = (100.0 * image_features @ text_features.T).softmax(dim=-1)
similarity = similarity.cpu().numpy() if self.device == "cuda" else similarity.numpy()
return similarity
except Exception as e:
self.logger.error(f"Error calculating similarity: {e}")
self.logger.error(traceback.format_exc())
raise
def preprocess_image(self, image: Union[Image.Image, np.ndarray]) -> torch.Tensor:
"""
預處理圖像以供CLIP模型使用
Args:
image: PIL圖像或numpy數組
Returns:
torch.Tensor: 預處理後的圖像張量
"""
try:
if not isinstance(image, Image.Image):
if isinstance(image, np.ndarray):
image = Image.fromarray(image)
else:
raise ValueError("Unsupported image format. Expected PIL Image or numpy array.")
image_input = self.preprocess(image).unsqueeze(0).to(self.device)
return image_input
except Exception as e:
self.logger.error(f"Error preprocessing image: {e}")
self.logger.error(traceback.format_exc())
raise
def process_image_region(self, image: Union[Image.Image, np.ndarray], box: List[float]) -> torch.Tensor:
"""
處理圖像的特定區域
Args:
image: 原始圖像
box: 邊界框 [x1, y1, x2, y2]
Returns:
torch.Tensor: 區域圖像的特徵
"""
try:
# 確保圖像是PIL格式
if not isinstance(image, Image.Image):
if isinstance(image, np.ndarray):
image = Image.fromarray(image)
else:
raise ValueError("Unsupported image format. Expected PIL Image or numpy array.")
# 裁剪區域
x1, y1, x2, y2 = map(int, box)
cropped_image = image.crop((x1, y1, x2, y2))
# 預處理並編碼
image_input = self.preprocess_image(cropped_image)
image_features = self.encode_image(image_input)
return image_features
except Exception as e:
self.logger.error(f"Error processing image region: {e}")
self.logger.error(traceback.format_exc())
raise
def batch_process_regions(self, image: Union[Image.Image, np.ndarray],
boxes: List[List[float]]) -> torch.Tensor:
"""
批量處理多個圖像區域
Args:
image: 原始圖像
boxes: 邊界框列表
Returns:
torch.Tensor: 所有區域的圖像特徵
"""
try:
# ensure PIL format
if not isinstance(image, Image.Image):
if isinstance(image, np.ndarray):
image = Image.fromarray(image)
else:
raise ValueError("Unsupported image format. Expected PIL Image or numpy array.")
if not boxes:
return torch.empty(0)
# 裁剪並預處理所有區域
cropped_inputs = []
for box in boxes:
x1, y1, x2, y2 = map(int, box)
cropped_image = image.crop((x1, y1, x2, y2))
processed_image = self.preprocess(cropped_image).unsqueeze(0)
cropped_inputs.append(processed_image)
# 批量處理
batch_tensor = torch.cat(cropped_inputs).to(self.device)
image_features = self.encode_image(batch_tensor)
return image_features
except Exception as e:
self.logger.error(f"Error batch processing regions: {e}")
self.logger.error(traceback.format_exc())
raise
def is_model_loaded(self) -> bool:
"""
檢查模型是否已成功載入
Returns:
bool: 模型載入狀態
"""
return self.model is not None and self.preprocess is not None
def get_device(self) -> str:
"""
獲取當前設備
Returns:
str: 設備名稱
"""
return self.device
def get_model_name(self) -> str:
"""
獲取模型名稱
Returns:
str: 模型名稱
"""
return self.model_name