Using AI and 3D Scanning to Recover Ancient Herculaneum Scrolls
利用 AI 與 3D 掃描還原古希臘赫庫蘭尼姆卷軸
Introduction
Researchers have successfully used high-resolution 3D scanning and artificial intelligence to read carbonized scrolls from the library of Herculaneum without needing to touch the fragile documents.
研究人員成功利用高解析度 3D 掃描與人工智慧，在無需接觸脆弱文件的情況下，閱讀了赫庫蘭尼姆圖書館的碳化卷軸。
Main Body
The Herculaneum scrolls were burned during the eruption of Mount Vesuvius in 79 AD. Since their discovery in 1752, they have been almost impossible to read because the charred papyrus is too fragile and the ink looks too similar to the blackened paper. In the past, trying to unroll these documents often caused them to break. However, the Vesuvius Challenge has reduced these risks by using particle accelerator scans and AI algorithms for 'virtual unwrapping.' Currently, these AI models are adjusted for each individual scroll, although the goal is to create a single system that can detect ink across the entire collection.
赫庫蘭尼姆卷軸在公元 79 年維蘇威火山爆發期間被燒毀。自 1752 年被發現以來，由於燒焦的莎草紙過於脆弱，且墨水顏色與變黑的紙張極為相似，幾乎無法閱讀。過去嘗試展開這些文件經常導致其破損。然而，「維蘇威挑戰賽」透過粒子加速器掃描與 AI 演算法進行「虛擬展開」，降低了這些風險。目前，這些 AI 模型是針對每卷卷軸個別調整的，但目標是建立一個單一系統，能夠偵測整個收藏集中的墨水。
Recently, researchers extracted 1.5 meters of text from a scroll dating back to the 2nd century BC. The text discusses ethics, art, and human nature, and it is likely that the Stoic philosopher Chrysippus wrote it. Furthermore, they recovered 70 columns from a work titled 'On Vices' and expanded another series by Philodemus. These discoveries are important because they provide original texts, which means historians no longer have to rely on summaries written by later authors that might be biased. The project is now moving from recovering small sections to reading entire scrolls.
最近，研究人員從一卷追溯至公元前 2 世紀的卷軸中提取了 1.5 公尺的文字。文本討論了倫理學、藝術與人性，很可能出自斯多葛哲學家克律西波斯之手。此外，他們還從一部名為《論惡習》的作品中還原了 70 欄文字，並擴充了菲洛得姆斯的另一系列作品。這些發現至關重要，因為它們提供了原典文本，這意味著歷史學家不再需要依賴後世作者可能帶有偏見的摘要。該計畫目前正從還原小片段過渡到閱讀整卷卷軸。
Conclusion
The combination of AI and advanced imaging has moved the project from a technical experiment to the systematic recovery of lost philosophical works.
AI 與先進影像技術的結合，使該計畫從一項技術實驗，轉變為對失傳哲學作品的系統性還原。
Vocabulary Learning
🚀 THE B2 LEAP: Moving from 'Simple' to 'Precise'
At the A2 level, you describe the world using basic verbs like get, do, make, or have. To reach B2, you must replace these with Precise Action Verbs.
Look at how the article transforms simple ideas into professional, academic English:
1. The "Get" Upgrade
- A2 approach: "They got the text from the scroll."
- B2 approach: "Researchers extracted 1.5 meters of text."
- Why it matters: "Extracted" implies a careful, technical process of removing something specific. It changes the tone from a casual conversation to a scientific report.
2. The "Change" Upgrade
- A2 approach: "The AI changed to fit each scroll."
- B2 approach: "AI models are adjusted for each individual scroll."
- Why it matters: "Adjusted" describes a small, precise change made to achieve a better result. This is a key B2 vocabulary shift: specifying how something changed.
3. The "Help" Upgrade
- A2 approach: "The AI helped the project move forward."
- B2 approach: "The combination of AI... has moved the project from a technical experiment to the systematic recovery..."
- Why it matters: Notice the use of "from [Point A] to [Point B]". B2 learners don't just say things improved; they describe the transition and the scale of the progress.
💡 Pro-Tip for your transition: Whenever you want to use the word do, get, or make, stop. Ask yourself: "What is the exact action happening here?"
- Is it recovering (finding something lost)?
- Is it expanding (making something larger)?
- Is it detecting (finding something hidden)?
Using these specific verbs is the fastest way to sound like a B2 speaker.