引导式测验 · 中等 等级

野生动物保护中的人工智能测验

了解保护人工智能如何分析相机陷阱、声学传感器、个人标记、巡逻历史和标记图像以支持监控。

相关引导路径人工智能在野生动物保护中的应用
问题 1 的 8

Wildlife Insights 记录的 AI 辅助功能如何处理红外触发相机拍摄的照片？

预览每个问题8 问题

当您参加测验时，就会出现答案和解释。

  1. Wildlife Insights 记录的 AI 辅助功能如何处理红外触发相机拍摄的照片？
    • Predicts species labels and can identify blank images for review
    • Guarantees identification of every animal in any ecosystem
    • Tracks individuals from satellite radar alone
    • Measures water quality from pipe vibrations
  2. 所引用的热带森林研究训练 CNN 检测什么声音事件？
    • Every kind of chainsaw and vehicle noise worldwide
    • Gunshots in acoustic recordings from a Belize forest study
    • Bird species in all habitats without training data
    • Water leaks in urban plastic pipes
  3. 在所引用的动物重新识别论文中，哪些视觉线索有助于区分有斑纹的个体？
    • GPS serial numbers attached to every animal
    • Barcodes implanted under the skin
    • Distinctive markings such as stripes, spots or other body patterns
    • The color of the habitat alone
  4. 在对偷猎者行为存在不确定性的情况下，PAWS 计算出什么类型的巡逻策略？
    • A single fixed route used every day
    • A route based only on animal migration distance
    • A path that guarantees no poaching incidents
    • A mixed strategy that makes patrol routes less predictable
  5. 为什么所引用的红外触发相机研究使用迁移学习进行物种识别？
    • It can reuse features learned from a larger related dataset when target labels are limited
    • It removes the need for training or validation images
    • It guarantees perfect accuracy for rare species
    • It replaces camera-trap images with GPS coordinates
  6. What did the cited thermal-drone study evaluate for nocturnal wildlife surveys?
    • Universal nighttime detection of every species and intruder
    • Thermal detection of white-tailed deer under the study’s warm-weather conditions
    • Species identification from satellite soil-moisture maps
    • Autonomous tracking of animal migration across continents
  7. Which historical data does the PAWS paper use to inform patrol planning?
    • Only tourist social-media posts
    • Only satellite pictures of clouds
    • Past patrol and poaching records from protected areas
    • Only DNA samples from captive animals
  8. How many species did the cited Snapshot Serengeti study train deep CNNs to identify, count and describe?
    • 4 species in 32 images
    • 480 species in a million satellite scenes
    • One species from 3.2 million audio files
    • 48 species in the 3.2-million-image dataset