BANG💥: Dividing 3D Assets via
Generative Exploded Dynamics
Longwen Zhang1,2 Qixuan Zhang1,2 Haoran Jiang1,2 Yinuo Bai1,2 Wei Yang3 Lan Xu1# Jingyi Yu1#
1ShanghaiTech University 2Deemos Technology 3Huazhong University of Science and Technology
SIGGRAPH 2025
👉ARXIV 👉DL.ACM 👉DEMO (coming soon)
Abstract
3D creation has always been a unique human strength, driven by our ability to deconstruct and reassemble objects using our eyes, mind and hand. However, current 3D design tools struggle to replicate this natural process, requiring considerable artistic expertise and manual labor. This paper introduces BANG, a novel generative approach that bridges 3D generation and reasoning, allowing for intuitive and flexible part-level decomposition of 3D objects. At the heart of BANG is "Generative Exploded Dynamics", which creates a smooth sequence of exploded states for an input geometry, progressively separating parts while preserving their geometric and semantic coherence.
BANG utilizes a pre-trained large-scale latent diffusion model, fine-tuned for exploded dynamics with a lightweight exploded view adapter, allowing precise control over the decomposition process. It also incorporates a temporal attention module to ensure smooth transitions and consistency across time. BANG enhances control with spatial prompts, such as bounding boxes and surface regions, enabling users to specify which parts to decompose and how. This interaction can be extended with multimodal models like GPT-4, enabling 2D-to-3D manipulations for more intuitive and creative workflows.
The capabilities of BANG extend to generating detailed part-level geometry, associating parts with functional descriptions, and facilitating component-aware 3D creation and manufacturing workflows. Additionally, BANG offers applications in 3D printing, where separable parts are generated for easy printing and reassembly. In essence, BANG enables seamless transformation from imaginative concepts to detailed 3D assets, offering a new perspective on creation that resonates with human intuition.
Overview
The overview illustrates the proposed framework for Generative Exploded Dynamics. The pipeline consists of four stages: Exploded Data Synthesis generates the training data. Exploded Dynamics Generator produces the exploded dynamics based on the input geometry. Trajectory Optimization refines the trajectories of the exploded parts, ensuring smooth reassembly of the components. Finally, Controllable Generation allows users to interactively control and refine the explosion by conditions.
Results
A fictional journey from Earth's surface to the far reaches of space, celebrating humanity's boundless ingenuity and spirit of discovery. Each object is generated from a concept image and illustrated in four assembly states, using parts generated from our Generative Exploded Dynamics.
A steampunk workshop, where blueprints transform into tangible reality, powered by our BANG framework. Each asset begins as a concept image generated by FLUX, is then transformed into an integral 3D mesh via our base generative model, and subsequently exploded into parts and are meticulously enhanced part-wise for maximum visual fidelity. The generated exploded structures are displayed against the backdrop, showcasing the enhanced details achieved through our exploded-enhance pipeline.
Citation
@article{zhang2025bang,
author = {Zhang, Longwen and Zhang, Qixuan and Jiang, Haoran and Bai, Yinuo and Yang, Wei and Xu, Lan and Yu, Jingyi},
title = {BANG: Dividing 3D Assets via Generative Exploded Dynamics},
year = {2025},
issue_date = {August 2025},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
volume = {44},
number = {4},
issn = {0730-0301},
url = {https://doi.org/10.1145/3730840},
doi = {10.1145/3730840},
abstract = {3D creation has always been a unique human strength, driven by our ability to deconstruct and reassemble objects using our eyes, mind and hand. However, current 3D design tools struggle to replicate this natural process, requiring considerable artistic expertise and manual labor. This paper introduces BANG, a novel generative approach that bridges 3D generation and reasoning, allowing for intuitive and flexible part-level decomposition of 3D objects. At the heart of BANG is "Generative Exploded Dynamics", which creates a smooth sequence of exploded states for an input geometry, progressively separating parts while preserving their geometric and semantic coherence. BANG utilizes a pre-trained large-scale latent diffusion model, fine-tuned for exploded dynamics with a lightweight exploded view adapter, allowing precise control over the decomposition process. It also incorporates a temporal attention module to ensure smooth transitions and consistency across time. BANG enhances control with spatial prompts, such as bounding boxes and surface regions, enabling users to specify which parts to decompose and how. This interaction can be extended with multimodal models like GPT-4, enabling 2D-to-3D manipulations for more intuitive and creative workflows. The capabilities of BANG extend to generating detailed part-level geometry, associating parts with functional descriptions, and facilitating component-aware 3D creation and manufacturing workflows. Additionally, BANG offers applications in 3D printing, where separable parts are generated for easy printing and reassembly. In essence, BANG enables seamless transformation from imaginative concepts to detailed 3D assets, offering a new perspective on creation that resonates with human intuition.},
journal = {ACM Trans. Graph.},
month = jul,
articleno = {62},
numpages = {21},
keywords = {generative exploded dynamics, partlevel 3D generation, 3D asset generation}
}
Notice: You may not copy, reproduce, distribute, publish, display, perform, modify, create derivative works, transmit, or in any way exploit any such content, nor may you distribute any part of this content over any network, including a local area network, sell or offer it for sale, or use such content to construct any kind of database.