clip

name	clip
description	Zero-shot image classification and image-text search.
version	1.0.0
author	Orchestra Research
license	MIT
dependencies	["transformers","torch","pillow"]
platforms	["linux","macos","windows"]
metadata	{"hermes":{"tags":["Multimodal","CLIP","Vision-Language","Zero-Shot","Image Classification","OpenAI","Image Search","Cross-Modal Retrieval","Content Moderation"]}}

CLIP —— 对比学习语言-图像预训练模型

OpenAI开发的能够通过自然语言理解图像的模型。

何时使用CLIP

适用场景：

• 零样本图像分类（无需训练数据）
• 图像与文本的相似度/匹配检测
• 语义图像搜索
• 内容审核（识别不当内容及暴力画面）
• 视觉问答
• 跨模态检索（图像→文本，文本→图像）

核心指标：

• GitHub星标数超过25,300个
• 基于4亿组图像-文本对进行训练
• 在ImageNet零样本任务上的表现可与ResNet-50相媲美
• 采用MIT许可证

可选替代方案：

• BLIP-2：更出色的图像描述功能
• LLaVA：视觉语言对话模型
• Segment Anything：图像分割工具

快速入门

安装指南

pip install git+https://github.com/openai/CLIP.git
pip install torch torchvision ftfy regex tqdm

零样本分类

import torch
import clip
from PIL import Image

# Load model
device = "cuda" if torch.cuda.is_available() else "cpu"
model, preprocess = clip.load("ViT-B/32", device=device)

# Load image
image = preprocess(Image.open("photo.jpg")).unsqueeze(0).to(device)

# Define possible labels
text = clip.tokenize(["a dog", "a cat", "a bird", "a car"]).to(device)

# Compute similarity
with torch.no_grad():
    image_features = model.encode_image(image)
    text_features = model.encode_text(text)

    # Cosine similarity
    logits_per_image, logits_per_text = model(image, text)
    probs = logits_per_image.softmax(dim=-1).cpu().numpy()

# Print results
labels = ["a dog", "a cat", "a bird", "a car"]
for label, prob in zip(labels, probs[0]):
    print(f"{label}: {prob:.2%}")

可用模型

# Models (sorted by size)
models = [
    "RN50",           # ResNet-50
    "RN101",          # ResNet-101
    "ViT-B/32",       # Vision Transformer (recommended)
    "ViT-B/16",       # Better quality, slower
    "ViT-L/14",       # Best quality, slowest
]

model, preprocess = clip.load("ViT-B/32")

模型	参数量	处理速度	质量表现
RN50	1.02亿	快	较好
ViT-B/32	1.51亿	中等	更优
ViT-L/14	4.28亿	慢	最佳

图像与文本相似度检测

# Compute embeddings
image_features = model.encode_image(image)
text_features = model.encode_text(text)

# Normalize
image_features /= image_features.norm(dim=-1, keepdim=True)
text_features /= text_features.norm(dim=-1, keepdim=True)

# Cosine similarity
similarity = (image_features @ text_features.T).item()
print(f"Similarity: {similarity:.4f}")

语义图像搜索

# Index images
image_paths = ["img1.jpg", "img2.jpg", "img3.jpg"]
image_embeddings = []

for img_path in image_paths:
    image = preprocess(Image.open(img_path)).unsqueeze(0).to(device)
    with torch.no_grad():
        embedding = model.encode_image(image)
        embedding /= embedding.norm(dim=-1, keepdim=True)
    image_embeddings.append(embedding)

image_embeddings = torch.cat(image_embeddings)

# Search with text query
query = "a sunset over the ocean"
text_input = clip.tokenize([query]).to(device)
with torch.no_grad():
    text_embedding = model.encode_text(text_input)
    text_embedding /= text_embedding.norm(dim=-1, keepdim=True)

# Find most similar images
similarities = (text_embedding @ image_embeddings.T).squeeze(0)
top_k = similarities.topk(3)

for idx, score in zip(top_k.indices, top_k.values):
    print(f"{image_paths[idx]}: {score:.3f}")

内容审核

# Define categories
categories = [
    "safe for work",
    "not safe for work",
    "violent content",
    "graphic content"
]

text = clip.tokenize(categories).to(device)

# Check image
with torch.no_grad():
    logits_per_image, _ = model(image, text)
    probs = logits_per_image.softmax(dim=-1)

# Get classification
max_idx = probs.argmax().item()
max_prob = probs[0, max_idx].item()

print(f"Category: {categories[max_idx]} ({max_prob:.2%})")

批量处理

# Process multiple images
images = [preprocess(Image.open(f"img{i}.jpg")) for i in range(10)]
images = torch.stack(images).to(device)

with torch.no_grad():
    image_features = model.encode_image(images)
    image_features /= image_features.norm(dim=-1, keepdim=True)

# Batch text
texts = ["a dog", "a cat", "a bird"]
text_tokens = clip.tokenize(texts).to(device)

with torch.no_grad():
    text_features = model.encode_text(text_tokens)
    text_features /= text_features.norm(dim=-1, keepdim=True)

# Similarity matrix (10 images × 3 texts)
similarities = image_features @ text_features.T
print(similarities.shape)  # (10, 3)

与向量数据库的集成

# Store CLIP embeddings in Chroma/FAISS
import chromadb

client = chromadb.Client()
collection = client.create_collection("image_embeddings")

# Add image embeddings
for img_path, embedding in zip(image_paths, image_embeddings):
    collection.add(
        embeddings=[embedding.cpu().numpy().tolist()],
        metadatas=[{"path": img_path}],
        ids=[img_path]
    )

# Query with text
query = "a sunset"
text_embedding = model.encode_text(clip.tokenize([query]))
results = collection.query(
    query_embeddings=[text_embedding.cpu().numpy().tolist()],
    n_results=5
)

最佳实践

1. 大多数场景下使用 ViT-B/32——性能与效率平衡良好
2. 对嵌入向量进行归一化处理——这是计算余弦相似度的必要步骤
3. 采用批量处理方式——效率更高
4. 缓存嵌入向量——重新计算成本极高
5. 使用描述性标签——能提升零样本任务的性能
6. 推荐使用 GPU——处理速度可提升 10 到 50 倍
7. 对图像进行预处理——请使用提供的预处理函数

性能表现

操作类型	CPU	GPU（V100）
图像编码	约 200 毫秒	约 20 毫秒
文本编码	约 50 毫秒	约 5 毫秒
相似度计算	小于 1 毫秒	小于 1 毫秒

局限性

1. 不适用于精细任务——更适合处理大类别场景
2. 需要描述性文本——模糊的标签会导致性能不佳
3. 基于网络数据存在偏差——可能存在数据集偏差
4. 不支持边界框识别——仅能处理整幅图像
5. 空间理解能力有限——对位置和数量的分析能力较弱

相关资源

• GitHub 仓库：https://github.com/openai/CLIP ⭐ 25,300+ 次点赞
• 论文链接：https://arxiv.org/abs/2103.00020
• Colab 实验环境：https://colab.research.google.com/github/openai/clip/
• 许可证：MIT 许可证