← Back to all news
arXiv AI By Andrea Ceron, Michael Schmidt, Alvaro Marcos-Ramiro, Sebastian Schmidt, Benjamin Busam

CRISP: Fixing Flying Pixels in Latent LiDAR Generation via Diffusion Decoding

Read the original on arXiv AI →

The paper introduces CRISP, a pixel‑space diffusion decoder designed to eliminate flying pixels in latent LiDAR generation. CRISP replaces existing video‑VAE and LiDAR‑native decoders while keeping the encoder and latent generator unchanged, and it incorporates a backbone‑agnostic latent adapter, a DiT‑based denoiser, and a support mask predictor. Across datasets such as KITTI‑360, SemanticKITTI, and nuScenes, CRISP reduces FSVD/FPVD by an average of 50.5% for frozen backbones and up to 71%/74% for generic video VAEs, achieving significant improvements in depth discontinuity handling and narrowing the sim‑to‑real gap in pretrained LiDM world models.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv Computer Vision
Oct 28

SpeeDiff: Scalable Pixel-Anchored End-to-End Latent Diffusion Model

SpeeDiff is a latent diffusion model that jointly trains a VAE and diffusion network from scratch, using a Tweedie Pixel Reconstruction loss to prevent latent collapse. The approach allows scaling a transformer‑based architecture and improves representation alignment. Experiments show the SpeeDiff‑XL model trains 140× faster than Vanilla SiT and 61× faster than REPA, achieving an FID of 1.50 on ImageNet 256×256 and 1.53 on 512×512 without guidance.

By Bingliang Zhang, Wenda Chu, Yizhuo Li, Linjie Yang, Yisong Yue, Katherine L. Bouman, Yang Song, Qiushan Guo
llmsdiffusionsafety
More like this →
arXiv Computer Vision
Oct 5

PatchScene: Patch-based Voxel Diffusion for Large-Scale Scene Completion

PatchScene is a diffusion-based framework that completes large-scale LiDAR scenes using a patch-based voxel diffusion paradigm, generating fine-grained geometry in localized 3D regions. It incorporates a confidence-guided spatio-temporal fusion mechanism to merge overlapping patches and adjacent frames, and introduces an Annular-Flow diffusion strategy that propagates high-fidelity information from near to far ranges. Experiments on SemanticKITTI show state-of-the-art performance, with the model trained on 20 m LiDAR ranges generalizing to 50 m scenes without retraining.

By Qingdong Xu, Jiajun Zhu, Shilin Zhu, Xinjing He, Chao Lu, Huanran Wang, Jiyao Zhang
agentsdiffusionroboticsbenchmarks
More like this →