應用機器學習進行海運貨櫃缺陷之自動偵測

 

產品分類：AI人工智慧與物聯網

廠商名稱：國立成功大學

攤位號碼：N008

 

產品特色

 

"Container Shipping Scale and Cost Impact
Container shipping transports 90% of the world's freight. Standardized design of containers enhances loading and unloading efficiently. Poor standards in cargo transport unit packing cost the transport and logistics industry around US$6 billion yearly. 2025 worldwide container transportation capacity is expected to reach 29.5 million TEUs."
"A CNN-Transformer Hybrid Framework
The framework includes three components: encoding, upsampling, and decoding. The encoding phase simulates missing parts by randomly deleting 60% of the original image. Upsampling enhances mask resolution, and the decoding phase predicts and fills in missing areas using a transformer-based self-supervised approach. Additionally, the CFE-VP module focuses on key local mask features to accelerate learning.

ML Platform Operated w/o Coding Knowledge
This study compares custom-coded CNN-Transformer deep learning models to Teachable Machine@, for autonomous maritime shipping container internal defect identification. Greatly reduces the technical barrier for non-expert users and SMEs using machine learning technology for practical applications. No-code platforms may not run optimal algorithm configurations. Many no-code platforms are cloud-based, therefore users' data must be processed on third-party servers. Firms that handle sensitive data may face security issues."

 

相關產品

 

• 

  以物理 AI 實現單一金屬氧化氣體感測器的多氣體分辨技術

• 

  國立成功大學AI Service Center

• 

  查詢式系統異常偵測技術

• 

  多維度應變感測裝置、藉由其的外部作用類型的感測方法以及三元值邏輯裝置的形成方法以及三元值邏輯系統

• 

  基於動態視窗演算法的機器人自主導航及噴灑系統

 

 

 

回列表