Викторина, связанная с руководством · Средний Уровень
Анализ уровня ошибок для викторины по криминалистике изображений
Проверьте остатки рекомпрессии JPEG, настройки качества, ошибки в истории обработки, ограничения на происхождение AI, сохранение исходного кода и осторожные заявления.
Вопрос 1 из 8
What does an ELA image primarily display?
Предварительный просмотр каждого вопроса8 вопросы
Ответы и пояснения появляются по мере прохождения теста.
- What does an ELA image primarily display?
- Pixel differences after a chosen JPEG recompression
- The true camera location
- A guaranteed edit mask
- Whether the scene occurred
- Why might an authentic image have an unusual ELA region?
- Every bright region is generated by AI
- Authentic JPEGs cannot be resaved
- A text overlay, high-detail texture or later processing may differ locally
- ELA ignores image content
- What does a uniform ELA result fail to rule out?
- That the file is a JPEG
- A manipulated or generated image saved with a consistent compression history
- That the tool used a quality setting
- That pixels were compared
- Which file is best to seek before interpreting ELA?
- Only a social-media screenshot
- Only an AI-generated thumbnail
- Only the smallest downloaded copy
- The earliest high-quality original available
- Why record recompression settings in an ELA report?
- They affect the visualization and let others reproduce it
- They prove the photographer’s identity
- They establish copyright ownership
- They replace source links
- What claim about a bright patch overstates ELA?
- It merits closer inspection
- It may have a different compression response
- It proves the patch is a pasted fake object
- It could reflect texture
- Which follow-up strengthens an image verification investigation?
- Rely on one ELA screenshot
- Check source history and independent contextual evidence
- Assume metadata is uneditable
- Ignore earlier copies
- Can ELA alone identify an image as AI-generated?
- Yes, if any area is bright
- Yes, if the JPEG is large
- Yes, if metadata are missing
- No; it measures compression behavior, not image origin