Context. Asteroids preserve important information about the origin and evolution of the Solar System. Three-dimensional (3D) surface models of asteroids are essential for exploration missions and scientific research. Regular methods for 3D surface reconstruction of asteroids, such as stereo-photogrammetry (SPG), usually struggle to reconstruct textureless areas and can only generate sparse surface models. Stereo-photoclinometry (SPC) can reconstruct pixel-wise topographic details but its performance depends on the availability of images obtained under different illumination conditions and suffers from uncertainties related to surface reflectance and albedo.Aims. This paper presents Asteroid-NeRF, a novel deep-learning method for 3D surface reconstruction of asteroids that is based on the state-of-the-art neural radiance field (NeRF) method.Methods. Asteroid-NeRF uses a signed distance field (SDF) to reconstruct a 3D surface model from multi-view posed images of an asteroid. In addition, Asteroid-NeRF incorporates appearance embedding to adapt to different illumination conditions and to maintain the geometric consistency of a reconstructed surface, allowing it to deal with the different solar angles and exposure conditions commonly seen in asteroid images. Moreover, Asteroid-NeRF incorporates multi-view photometric consistency to constrain the SDF, enabling optimised reconstruction.Results. Experimental evaluations using actual images of asteroids Itokawa and Bennu demonstrate the promising performance of Asteroid-NeRF, complementing traditional methods such as SPG and SPC. Furthermore, due to the global consistency and pixel-wise training of Asteroid-NeRF, it produces highly detailed surface reconstructions. Asteroid-NeRF offers a new and effective solution for high-resolution 3D surface reconstruction of asteroids that will aid future exploratory missions and scientific research on asteroids.

Multimodal fusion unlocks the potential of point cloud semantic segmentation, thereby driving advancements in surface observation and visual perception tasks. Although light detection and ranging (LiDAR) systems capture precise 3D structural geometry and optical images provide rich semantic and textural information, existing fusion methods struggle with limited cross-modal perception and insufficient information complementarity. To address these limitations, we propose a multi-stage LiDAR-image collaborative perception fusion network (MCPFNet) for point cloud semantic segmentation of urban scenes. At the middle fusion stage, the network incorporates an elevation-guided geometric-aware fusion module and a semantic-aware cross-attention fusion module to enable bidirectional feature injection between LiDAR and image modalities. In the late fusion stage, a bidirectional adaptive fusion module further refines semantic representations through gated weighting and bidirectional cross-attention mechanisms. Extensive experiments on three multimodal datasets with different resolutions, i.e., ISPRS Vaihingen, N3C-California, and UAVScenes, demonstrate that MCPFNet outperforms existing fusion methods, achieving mIoUs of 74.51%, 95.15%, and 62.76%, respectively. Hence, our multi-stage fusion and bidirectional interaction strategy is more reliable and accurate than existing methods in performing segmentation across diverse and complex urban scenes.

Transient distractors in dynamic, real-world 3D scenes challenge 3D Gaussian splatting (3DGS) by violating the multi-view consistency assumption, which is essential for accurate 3D reconstruction. Existing methods struggle to accurately identify the boundaries, shadows, and small-scale regions of transient distractors, causing residual transient artifacts to remain in the scenes. Meanwhile, valid static regions may be incorrectly identified as transient distractors, resulting in the loss of useful scene observations and further degrading reconstruction quality. To address these challenges, we propose DifSplat, a novel distractor-free 3D scene reconstruction framework that exploits the priors of visual foundation models. Specifically, we combine an adaptive feature-aware uncertainty estimation strategy with a transient mask optimization strategy guided by high-resolution image residuals, enabling more accurate identification of transient distractors while preserving the static scene. Furthermore, we introduce an inpainting-guided image replacement strategy that utilizes a video diffusion model to inpaint only the regions occluded by transient distractors in the renderings. The inpainted regions are then used to replace the corresponding regions in the original input views, providing additional static scene cues for fine-tuning 3DGS. Extensive experiments on benchmark datasets demonstrate that our proposed method achieves state-of-the-art distractor-free reconstruction quality in both indoor and complex outdoor scenes.