圖像描述/影像自動語義生成 (Image Captioning)看圖說故事_V2L問題(Visual-to-Language)

 

https://www.oreilly.com/library/view/deep-learning-for/9781788295628/89def52b-a455-4a2f-b51e-23b74e154bd0.xhtml

Image captioning is the task of describing the image with text

圖像描述主要應用十分多領域

• 從醫學影像產生臨床報告
• 旅遊照之情感分析與自動撰寫評論
• 影片摘要
• 視覺問答

國內各大院校研究論文

國立交通大學-資訊學院資訊學程/深度學習–旅遊照之情感分析與自動撰寫評論(2019)

https://hdl.handle.net/11296/mh28rj

國立中山大學-電機工程學系研究所/基於 Transformer 具領域外泛化能力之影像標題生成(2021)

https://hdl.handle.net/11296/mmenf5

國立成功大學-資訊工程學系/基於模態轉換和大型語言模型的視覺問答(2023)

https://hdl.handle.net/11296/95bdq4

大同大學/資訊工程學系/利用數種深度學習搭配注意力機制對胸腔X光照做醫療報告生成(2023)

https://hdl.handle.net/11296/55b3js

朝陽科技大學-營建工程系/工地影像字幕生成技術初探-以工地危害描述為例(2024)

https://hdl.handle.net/11296/4jhbej

國立臺灣科技大學-工業管理系/結合圖片描述技術與影片標題於影片摘要預測之應用(2024)

https://hdl.handle.net/11296/zxz842

國立清華大學-資訊系統與應用研究所/使用基於圖形的深度轉換器與大型語言模型來從醫學影像產生臨床報告(2024)

https://hdl.handle.net/11296/az53h4

Image Captioning in news report scenario

https://arxiv.org/abs/2403.16209

Mitigating Gender Bias in Natural Language Processing: Literature Review

https://arxiv.org/abs/1906.08976

Patch Matters: Training-free Fine-grained Image Caption Enhancement via Local Perception

https://arxiv.org/abs/2504.06666

Deep Learning Image Captioning Technology for Business Applications

https://www.iotforall.com/deep-learning-image-captioning-technology-for-business-applications

Automatic image captioning in Thai for house defect using a deep learning-based approach

(運用深度學習方法自動生成泰語房屋缺陷影像說明)

https://link.springer.com/article/10.1007/s43674-023-00068-w

https://github.com/manadda-j/deep-learning

A deep learning-based image captioning method to automatically generate comprehensive explanations of bridge damage(一種基於深度學習的影像標註方法，可自動生成橋樑損壞的全面解釋)

https://onlinelibrary.wiley.com/doi/epdf/10.1111/mice.12793

https://onlinelibrary.wiley.com/doi/full/10.1111/mice.12793

Empirical Study of Image Captioning Models Using Various Deep Learning Encoders

https://link.springer.com/chapter/10.1007/978-981-99-0047-3_27?fromPaywallRec=true

Experimenting Encoder-Decoder Architecture for Visual Image Captioning

https://link.springer.com/chapter/10.1007/978-3-031-22405-8_16?fromPaywallRec=true

Image Captioning with Multiple Perspectives—A Visual Context-Based Approach

https://link.springer.com/chapter/10.1007/978-981-97-4711-5_30?fromPaywallRec=true

Google相簿自動辨識技術出大包，竟把用戶的黑人朋友標示成大猩猩！

https://www.techbang.com/posts/24479-google-photo-gorilla

相片辨識出包誤將黑人標成大猩猩，Google火速道歉

https://www.ithome.com.tw/news/97131

將黑人標註為「靈長類動物」 臉書AI功能出包急道歉

https://news.ltn.com.tw/news/life/breakingnews/3656318?utm_source=NEWS&utm_medium=1&utm_campaign=MOREPAGE

Google AI 將黑人識別成「大猩猩」兩年後：就算能識別，也不敢識別了

https://inboundnow.org/topics/38474/google-ai-%E5%B0%87%E9%BB%91%E4%BA%BA%E8%AD%98%E5%88%A5%E6%88%90%E3%80%8C%E5%A4%A7%E7%8C%A9%E7%8C%A9%E3%80%8D%E5%85%A9%E5%B9%B4%E5%BE%8C%EF%BC%9A%E5%B0%B1%E7%AE%97%E8%83%BD%E8%AD%98%E5%88%A5%EF%BC%8C/