Moonchu China Foto Collection
2022 Ongoing | Visualization Research Centre, HKBU
Section 1: Project Overview
The Moonchu China Foto Collection invites visitors to explore the historical landscapes of old China through the lens of explorer and photographer James Ricalton (New York, 1844). Featuring a carefully selected subset of 100 photographs from his collection, along with stories from his book China Through the Stereoscope (1901), this project brings the past to life in a 1:1 scale, immersive 3D experience. The aesthetic qualities of Ricalton’s stereographic plates are digitally enhanced, amplifying the visual and contextual depth of these images, which offer a vivid glimpse into China’s history.
Section 2: Technical Narrative
This immersive experience is supported by a comprehensive Narrative Data Repository, where the digitized photographs and Ricalton’s narratives are stored. The repository contextualizes each photograph by connecting it with relevant narratives, enabling a seamless search and personalized interaction. Using advanced language model techniques, the platform aligns narratives with user input, guiding users toward specific photographs and providing customized answers to their questions.
Further, the platform’s entity recognition technology identifies connections within the texts, facilitating easy exploration by allowing users to filter content based on themes or key figures. Additionally, a gaze-tracking feature enables collaborative photo colorization, inviting visitors to deepen their engagement with these historical images in a shared 3D environment.
第一部分：項目概述
《Moonchu 中國照片收藏》讓訪客透過探險家兼攝影師詹姆斯·里卡爾頓（James Ricalton, 1844年，紐約）的鏡頭，探索舊中國的歷史風貌。該項目精選了他的收藏中100張照片，並結合他1901年出版的《中國立體視圖》書中的故事，以1:1比例的沉浸式3D體驗呈現出過去的場景。里卡爾頓的立體照片經過數位增強，使視覺效果和歷史背景更具深度，生動地再現了中國歷史的縮影。
第二部分：技術敘述
這項沉浸式體驗由一個全面的敘事數據庫支撐，裡面保存了數位化的照片和里卡爾頓的敘事內容。這個數據庫將每張照片與相關敘事相連，實現無縫的搜尋和個性化互動體驗。透過先進的語言模型技術，該平台能夠根據使用者的輸入匹配敘事，引導使用者找到特定照片，並針對選定的影像提供定製化的回應。
此外，該平台的實體識別技術能夠在文本中辨識出敘事間的聯繫，使使用者可根據主題或關鍵人物進行篩選探索。同時，注視追蹤功能讓訪客可以共同為照片上色，在這個共享的3D環境中加深對這些歷史影像的互動體驗。
Chief Investigator: Jeffrey SHAW
Institution: …
Date: 2023 - ongoing
Team: Renjie MAN (image colorization), Susanne POON (image scanning), Jeremy YIP (image colorization), Jiaming LIAO (image scanning), Paul BETHGE (programming)
Hardware: nVis with 29.6 surround sound system, Cluster System
Software: Unreal Engine
Johnson CHEUNG
The China Academy of Art
Innovation and Technology Fund, HKSAR
