Spaces:
Sleeping
Sleeping
File size: 4,603 Bytes
8f4216f 863a20f 421c641 8f4216f 6dca484 0dc317f 6dca484 bd43f58 6dca484 863a20f 6dca484 8f4216f 6dca484 863a20f 6dca484 863a20f 6dca484 8f4216f | 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 | # -*- coding: utf-8 -*-
"""video.ipynb
Automatically generated by Colab.
Original file is located at
https://colab.research.google.com/drive/1d-uwY0B5q7gOItN6fyA-1RoBcDwkOxb6
"""
# ✅ 全局加载模型一次（推荐）
from faster_whisper import WhisperModel
whisper_model = WhisperModel("tiny", device="cpu", compute_type="int8")
# 将整段video.py封装为analyze_video函数
def analyze_video(file, lang=None):
output = "" # 用于收集结果文本
import os
import zipfile
from dotenv import load_dotenv
from ultralytics import YOLO
import cv2
import requests
import base64
import json
# === 配置 ===
VIDEO_PATH = file
YOLO_MODEL_PATH = "trained_dataset/runs/detect/train/weights/best.pt"
# ✅ 自动解压模型文件（如果还没解压）
if not os.path.exists("trained_dataset"):
print("📦 正在解压 YOLO 模型文件...")
with zipfile.ZipFile("trained_dataset.zip", "r") as zip_ref:
zip_ref.extractall(".")
print("✅ 解压完成")
# ✅ 加载环境变量
from dotenv import load_dotenv
load_dotenv()
API_KEY = os.getenv("GOOGLE_VISION_API_KEY")
VISION_API_URL = f"https://vision.googleapis.com/v1/images:annotate?key={API_KEY}"
LIKELIHOOD_MAPPING = {"UNKNOWN": 0, "VERY_UNLIKELY": 1, "UNLIKELY": 2, "POSSIBLE": 3, "LIKELY": 4, "VERY_LIKELY": 5}
# === 初始化 ===
model = YOLO(YOLO_MODEL_PATH)
cap = cv2.VideoCapture(VIDEO_PATH)
safe_search_results = {"adult": 0, "spoof": 0, "medical": 0, "violence": 0, "racy": 0}
detected_texts = set()
detected_labels = set()
total_frames = 0
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break
total_frames += 1
# YOLO 检测
results = model.predict(source=frame, conf=0.2, save=False, verbose=False)
if len(results[0].boxes) > 0:
for c in results[0].boxes.cls:
class_id = int(c.item())
class_name = results[0].names[class_id]
detected_labels.add(class_name)
# 每 10 帧 Google 内容分析
if total_frames % 10 == 0:
_, buffer = cv2.imencode('.jpg', frame)
img_base64 = base64.b64encode(buffer).decode()
payload = {
"requests": [{
"image": {"content": img_base64},
"features": [
{"type": "SAFE_SEARCH_DETECTION"},
{"type": "TEXT_DETECTION"}
]
}]
}
response = requests.post(VISION_API_URL, json=payload)
result = response.json()
if "responses" in result and len(result["responses"]) > 0:
safe = result["responses"][0].get("safeSearchAnnotation", {})
for key in safe_search_results.keys():
if LIKELIHOOD_MAPPING.get(safe.get(key, "UNKNOWN"), 0) >= 4:
safe_search_results[key] += 1
texts = result["responses"][0].get("textAnnotations", [])
for text in texts:
detected_texts.add(text["description"])
cap.release()
# === Whisper 音频转文字 ===
output += "\n🎙 正在转录音频文本...:\n"
segments, info = whisper_model.transcribe(VIDEO_PATH, language=lang if lang != "auto" else None)
transcribed_text = "".join([seg.text for seg in segments])
# === 输出结果 ===
output += "\n🎯 YOLO检测到的图案类别:\n"
if detected_labels:
output += "✅ 检测到图案：" + "、".join(detected_labels) + "\n"
else:
output += "❌ 未检测到任何图案\n"
def risk_level(count, total):
if count > total * 0.05:
return "⛔️ 高风险"
elif count > 5:
return "⚠️ 中等风险"
else:
return "✅ 低风险"
output += "\n🔎 内容安全风险分析（每类满足可能性 ≥ LIKELY 的帧计数）:\n"
for k, v in safe_search_results.items():
output += f"{k.capitalize():<10}: {risk_level(v, total_frames)}（{v} 帧）\n"
output += "\n📝 视觉文字识别（OCR）:\n"
output += " ".join(detected_texts) + "\n" if detected_texts else "无可识别文字\n"
output += "\n🔊 Whisper语音识别结果:\n"
output += transcribed_text + "\n" if transcribed_text.strip() else "无有效语音\n"
return output
|