Position Forcing: Self-Conditioning 3D Generation

Ziheng Ouyang; Zeqiang Lai; Jiarui Chen; Jiangshan Wang; Yuhao Wan; Jingbo Gong; Xiangyu Yue; Hengshuang Zhao

cs.CVOct 7, 2026

Position Forcing: Self-Conditioning 3D Generation

Authors: Ziheng Ouyang, Zeqiang Lai, Jiarui Chen, Jiangshan Wang, Yuhao Wan, Jingbo Gong, Xiangyu Yue, Hengshuang Zhao, +2 more

Organizations: VCIP, Nankai University · Tencent Hunyuan · MMLab, CUHK · Fudan University · Shanghai Innovation Institute · HKU

Abstract

Recent single-stage 3D generative models commonly adopt VecSet representations, encoding 3D shapes as unordered sets of latent tokens. However, compared with two-stage methods that provide explicit positional guidance, these models must implicitly infer token positions throughout denoising, limiting their generation quality. We observe that, despite the absence of explicit positional conditioning, VecSet tokens retain recoverable spatial correspondences. Building on this observation, we propose Position Forcing, a position-based self-conditioning framework. During denoising, Position Forcing recovers token positions from the current clean latent estimate, quantizes them at progressively finer resolutions according to the denoising stage, and feeds the resulting positional encodings back into the diffusion Transformer. This progressively refined positional feedback provides spatial guidance at a granularity appropriate to each denoising stage, guiding shape generation along a coarse-to-fine trajectory and substantially improving generation quality without a separate position generation stage. Experiments demonstrate that Position Forcing achieves strong performance among single-stage 3D generative methods and outperforms several competitive multi-stage approaches.

Figures & tables

Figure 1 : Our observation and positional conditioning designs. (I) Latents can be decoded into shape geometry and token positions. (II) (a) VecSet uses no explicit positional conditioning; (b) VoxSet and two-stage pipelines use positions established beforehand. (c) Our naive design recovers positions from the current noisy state. (d) Position Forcing recovers positions from the predicted clean state and applies progressive quantization to provide coarse-to-fine spatial guidance.

Figure 2 : Method overview. (a) Position VAE learns latent tokens that preserve geometry–position correspondences, enabling both shape geometry and token-associated spatial positions to be decoded. (b) During denoising, Position-Forced DiT recovers token positions from the predicted clean latents and injects them as positional guidance into subsequent denoising steps through RoPE. (c) The recovered positions are progressively quantized from coarse to fine as denoising proceeds.

	Noise Level (%)
Training Strategy	0	10	20	30
Frozen VAE	67.4	51.3	23.0	6.7
Joint Finetuning	99.9	99.7	92.5	67.3

Table 1 : Effect of joint training on position recovery. Accuracy (%) measures the fraction of predicted positions that fall into the same cells as their corresponding query positions on a

128^{3}

grid.

Method	Latent Size	CD $\downarrow$	F1 $\uparrow$
TripoSG	$64\times 4096$	17.96	92.33
[ Li et al., 2025b ]	$64\times 8192$	16.61	94.19
Hunyuan3D-2.1	$64\times 4096$	9.08	89.43
[ Hunyuan3D et al., 2025 ]	$64\times 8192$	8.28	90.99
	$64\times 20480$	7.62	92.06
Position Forcing	$64\times 4096$	8.28	90.74

Table 2 : Quantitative comparisons of geometry reconstruction.

Model	Single Stage	ULIP-T $\uparrow$	ULIP-I $\uparrow$	Uni3D-T $\uparrow$	Uni3D-I $\uparrow$
TRELLIS [ Xiang et al., 2025 ]	✗	0.076	0.126	0.249	0.311
TRELLIS 2 [ Xiang et al., 2026 ]	✗	0.077	0.124	0.245	0.317
Direct3D-S2 [ Wu et al., 2026b ]	✗	0.074	0.122	0.247	0.314
Hi3DGen [ Ye et al., 2025 ]	✗	0.065	0.113	0.252	0.301
CraftsMan 1.5 [ Li et al., 2024 ]	✓	0.074	0.129	0.237	0.298
UniLat3D [ Wu et al., 2026a ]	✓	0.071	0.119	0.252	0.307

Table 3 : Comparison with existing 3D generation methods. Best results are highlighted in bold.

Figure 3 : Qualitative comparison of image-conditioned 3D generation. The input images are shown in the top row. Compared with existing methods, our approach produces more faithful 3D geometry with coherent global structures and fine-grained details across diverse object categories. Methods marked with * adopt a two-stage generation framework.

Figure 4 : Qualitative and quantitative ablation of Position Forcing. Columns (a)–(g) share the same configurations across qualitative and quantitative results. Query, Pred, and $x_{0}$ denote ground-truth query positions, predicted positions, and positions recovered from the predicted clean latent, respectively. Prog. denotes progressive position quantization. Configuration (g) uses the Stage-I VAE instead of the jointly trained VAE. Bold and underlining denote the best and second-best results.

Figure 5 : Positional consistency during sampling. We compare cell accuracy along the sampling trajectory for positions recovered from predicted clean latents (

P_{0\mid t}

) and noisy latents (

P_{t}

). Accuracy is the percentage of tokens whose intermediate and final decoded positions fall into the same cell, using progressive resolution

R(t)

or fixed resolution

R_{\max}

for both.

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 6 : Generation trajectory visualization. We use a fixed 50-step sampling schedule and visualize the predicted clean sample $\hat{x}_{0}$ at different generation steps. The intermediate predictions show that the model progressively forms the global structure in the early steps and then refines local details in later steps, eventually converging to the final output.

Figure 7 : Additional generation results. Generated geometries are shown with their corresponding input images inset at the bottom right.

Figure 8 : Multi-view qualitative comparison of image-conditioned 3D generation. Position Forcing produces smoother surfaces and more geometrically plausible details, particularly in rear views, with fewer surface artifacts and structural distortions. Methods marked with * adopt a two-stage generation framework.

Explore similar work

SuperVoxelGPT: Adaptive and Ordered 3D Tokenization for Autoregressive Shape Generation
May 28, 2026Yuan Li, Congyi Zhang, Xifeng Gao +1Voxel Autoregressive Generation
University of Texas at Dallas, USA · LightSpeed, USA
Flow3D-OPD: Multi-Teacher On-Policy Distillation for 3D Geometry Generation with Flow-Matching Diffusion Transformer
Sep 7, 2026Zhiwei Ning, Zhen Zhou, Puhua Jiang +73D Generative Models Uni-Opd
GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation
Mar 27, 2026Nicolas von Lützow, Barbara Rössle, Katharina Schmid +13D Scene Generation 3D Gaussian
Technical University of Munich, Germany

May 28, 2026, cs.CVJ/K move · Enter open · S save

SuperVoxelGPT: Adaptive and Ordered 3D Tokenization for Autoregressive Shape Generation

Yuan Li, Congyi Zhang, Xifeng Gao, Xiaohu Guo

University of Texas at Dallas, USA · LightSpeed, USA

Voxel Autoregressive Generation Visual Generation Autoregressive Language Models

Open paper AskPDF arXiv

Autoregressive multimodal large language models (MLLMs) enable 3D generation but struggle to scale to high-resolution shapes due to inadequate 3D tokenizations. Compact set-based representations discard deterministic spatial ordering, leading to ambiguous sequence prediction, while uniform or octree-based voxel grids preserve ordering at the cost of severe redundancy and excessively long sequences. This structural trade-off limits stable and efficient autoregressive 3D generation. We present SuperVoxelGPT, a representation-first framework that resolves this tension through adaptive and deterministically ordered supervoxel tokenization. Given a prompt, we first predict a coarse geometric saliency distribution and construct a shape-adaptive supervoxel partition using saliency-guided centroidal Voronoi tessellation, allocating fine-grained cells to complex regions and larger cells to smooth regions. Conditioned on this prompt and ordered supervoxel layout, we introduce a SuperVoxelVAE and fine-tune a pretrained MLLM to autoregressively generate supervoxel tokens. Experiments using Trellis-500K data show that SuperVoxelGPT reduces token sequence length to 12.8% of uniform voxel tokenization while achieving state-of-the-art generation quality and an average 10x speedup over prior methods.

Abstract

Explore similar work

SuperVoxelGPT: Adaptive and Ordered 3D Tokenization for Autoregressive Shape Generation

Flow3D-OPD: Multi-Teacher On-Policy Distillation for 3D Geometry Generation with Flow-Matching Diffusion Transformer

GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation