Don't wanna be here? Send us removal request.
Text
Subgraph Structure Extraction
Segmenting point cloud substructures through spatial clustering.
Next step: adopt different generation strategies for different types of structures.
0 notes
Text
A new point cloud reconstruction method

Connect neighboring points in the point cloud to form a network. When two edges share a common node, a triangle face is created.
0 notes
Text
Deformation Animation Update: The 3D models now use the original mesh surfaces to preserve more details. The original model size is tripled, while the expansion factor is reduced to one-third, making the original geometry more visually clear. Deformation is now time-controlled: it will cease 200 frames after the model stops moving.
0 notes
Text
Set a duration for the movement of the object. Prevent objects from undergoing endless movement or deformation.
0 notes
Text



Replace the point cloud with a grid surface model to retain more details
0 notes
Text
DISTURBANCE FIELD
This project investigates how narrative can intervene in the generative structure of urban space. By extracting structural language from contemporary news and ancient Greek drama, it constructs a text-based spatial mapping mechanism. Adopting a structuralist perspective, it decodes the relational networks embedded in language and projects them onto the city. Events, characters, and actions materialize into spatial movements and transformations—3D entities respond to narrative impulses as they move, mutate, and recombine across the urban fabric. The city thus becomes a dramatized spatial theater.
0 notes
Text
Text Dataset
News dataset: 9,908 news articles from Athens, collected from 2022 to 2024.
Drama dataset: 48 surviving ancient Greek tragedies, used as a structured narrative reference.
0 notes
Text
Narrative Structural Vectors: Defining Event Roles in Language
0 notes
Text
Narrative Structure Field: Revealing the City’s Latent Syntax
0 notes
Text
Mapping Ancient Drama into Contemporary Urban Space
Match the five-dimensional vectors of news events and dramatic events using Euclidean distance to obtain spatial embedding coordinates for the dramatic texts. Then apply spatial clustering to identify potential design sites.
0 notes
Text
Narrative Tension Field
We calculate the narrative tension of each dramatic event as the difference between its total out-degree and in-degree. These values serve as sources of spatial disturbance, and are interpolated using inverse distance weighting to generate the narrative tension field.
0 notes