Flux 2 provides superior precision for image creation, particularly when the process involves transformation, integration, or composition. Being free and open-source, it stands as a compelling alternative to Nano banana.
In architecture, certain functionalities are invaluable for rapidly transforming concepts into visual representations.
Unlike a standard VAE, this approach utilizes novel nodes: an Inpaint crop acting as a ‘spying glass’ and the Flux Enhancer node, which is instrumental in preventing the model from generating hallucinations.
Gaussian Splatting is a novel 3D scene representation technique that’s revolutionizing novel view synthesis (creating new camera viewpoints of 3D scenes). It emerged as a state-of-the-art alternative to Neural Radiance Fields (NeRFs) in 2023.
Core Concept
Instead of representing scenes with neural networks (like NeRFs) or explicit meshes, Gaussian Splatting represents scenes as millions of tiny 3D Gaussian « blobs » that can be efficiently rendered using a technique called « splatting. »
The technique represents a significant shift from implicit neural representations back toward explicit, differentiable graphics primitives, but with modern optimization techniques that achieve unprecedented quality and speed.
Before exploring and analyzing the corpus chosen for the representation of atmospheres, a simple example will be observed to measure the difficulty of the game of the representation of atmospheres in the project. This example is that of the choice that a designer could have between the strip window and the vertical window of the Renaissance type. This example has already been mentioned in the previous chapter as representative of the gap that can occur between an architectural discourse, the realization and the perception of the users. It seemed interesting to us at the beginning of the discussion on figuration to show that representation is a complex act passing through words and drawing.
