# FPSGen: Flexible Point Cloud Scene Generation with BEV-Supported Transport Flows

arXiv: https://arxiv.org/abs/2607.26645 · HTML: https://mlsanity.com/paper/2607.26645/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 36 of its 49 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Builds on (most cited first, 30 of 36)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow](https://mlsanity.com/paper/2209.03003/) (2022)
- [PointPillars: Fast Encoders for Object Detection from Point Clouds](https://mlsanity.com/paper/1812.05784/) (2018)
- [4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks](https://mlsanity.com/paper/1904.08755/) (2019)
- [Diffusion Probabilistic Models for 3D Point Cloud Generation](https://mlsanity.com/paper/2103.01458/) (2021)
- [LION: Latent Point Diffusion Models for 3D Shape Generation](https://mlsanity.com/paper/2210.06978/) (2022)
- [3D Shape Generation and Completion through Point-Voxel Diffusion](https://mlsanity.com/paper/2104.03670/) (2021)
- [PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows](https://mlsanity.com/paper/1906.12320/) (2019)
- [Semantic Scene Completion from a Single Depth Image](https://mlsanity.com/paper/1611.08974/) (2016)
- [PCN: Point Completion Network](https://mlsanity.com/paper/1808.00671/) (2018)
- [VoxFormer: Sparse Voxel Transformer for Camera-based 3D Semantic Scene Completion](https://mlsanity.com/paper/2302.12251/) (2023)
- [PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformers](https://mlsanity.com/paper/2108.08839/) (2021)
- [LMSCNet: Lightweight Multiscale 3D Semantic Completion](https://mlsanity.com/paper/2008.10559/) (2020)
- [GaussianFormer: Scene as Gaussians for Vision-Based 3D Semantic Occupancy Prediction](https://mlsanity.com/paper/2405.17429/) (2024)
- [Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completion](https://mlsanity.com/paper/2111.12702/) (2021)
- [UniScene: Unified Occupancy-centric Driving Scene Generation](https://mlsanity.com/paper/2412.05435/) (2024)
- [A Conditional Point Diffusion-Refinement Paradigm for 3D Point Cloud Completion](https://mlsanity.com/paper/2112.03530/) (2021)
- [GaussianFormer-2: Probabilistic Gaussian Superposition for Efficient 3D Occupancy Prediction](https://mlsanity.com/paper/2412.04384/) (2024)
- [Fast Point Cloud Generation with Straight Flows](https://mlsanity.com/paper/2212.01747/) (2022)
- [OPUS: Occupancy Prediction Using a Sparse Set](https://mlsanity.com/paper/2409.09350/) (2024)
- [SemCity: Semantic Scene Generation with Triplane Diffusion](https://mlsanity.com/paper/2403.07773/) (2024)
- [Pyramid Diffusion for Fine 3D Large Scene Generation](https://mlsanity.com/paper/2311.12085/) (2023)
- [Scaling Diffusion Models to Real-World 3D LiDAR Scene Completion](https://mlsanity.com/paper/2403.13470/) (2024)
- [Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer](https://mlsanity.com/paper/2202.09367/) (2022)
- [LODE: Locally Conditioned Eikonal Implicit Scene Completion from Sparse LiDAR](https://mlsanity.com/paper/2302.14052/) (2023)
- [Not-So-Optimal Transport Flows for 3D Point Cloud Generation](https://mlsanity.com/paper/2502.12456/) (2025)
- [Distilling Diffusion Models to Efficient 3D LiDAR Scene Completion](https://mlsanity.com/paper/2412.03515/) (2024)
- [LiDPM: Rethinking Point Diffusion for Lidar Scene Completion](https://mlsanity.com/paper/2504.17791/) (2025)
- [PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction](https://mlsanity.com/paper/2510.05613/) (2025)

## Abstract (arXiv)

Existing point-based generative methods for outdoor scenes primarily focus on LiDAR-conditioned completion. During training, noisy point clouds are constructed by perturbing complete ground-truth scenes, whereas during inference, they are initialized by adding noise to duplicated partial scans. This train-inference mismatch inherits the sparsity and visibility bias of partial scans, leading to sparse distant regions and incomplete geometry in occluded areas. Moreover, the reliance on partial scans restricts generation when LiDAR observations are unavailable or replaced by layout cues. We present FPSGen, a flexible framework that constructs point sources independently of partial scans. FPSGen first predicts a bird's-eye-view (BEV) prior with density, height, and mask channels from the active cues. The density map is then sampled to form a BEV-supported point source, enabling both unconditional and conditioned initialization. A teacher-student approximate optimal transport scheme then uses teacher-predicted endpoints to learn a velocity field that induces straighter transport paths. By integrating BEV point source construction with path-straightening transport, FPSGen provides a unified framework for unconditional and flexible cue-conditioned scene generation. Extensive experiments show that FPSGen achieves state-of-the-art JSD and voxel IoU performance on SemanticKITTI completion while maintaining strong performance with a single point transport step. On KITTI-360 unconditional generation, it also achieves the best Coverage (COV) among the compared methods.