LiDAR-Supervised Monocular Depth Estimation via Cross-Modal Supervision

DXPR: Depth-Based Vision-LiDAR Cross-Modal Place Recognition Using Vision Foundation Models

We present DXPR, a depth-based cross-modal place recognition (CMPR) framework that uses vision foundation models (VFMs) to match monocular camera queries against a LiDAR map without modality-specific encoders. This enables robots and autonomous vehicles to robustly localize using only cameras within pre-built LiDAR maps, even under severe seasonal, weather, and illumination changes. The key idea is to convert both camera images and LiDAR scans into a unified depth image representation so that a single VFM backbone with an aggregation head can learn modality-invariant global descriptors. To make pairwise metric learning faithful to scene geometry, we introduce a geometry-aware overlap miner: after cross-modal scale alignment of camera and LiDAR depth, we forward-warp measurements between views to compute a pixel-level overlap score. This score relabels ambiguous pairs and adaptively modulates the positive margin in a multi-similarity loss to avoid overfitting on weakly overlapping views. Extensive experiments on KITTI odometry and Boreas demonstrate strong performance and robustness across seasons, weather, and day/night. On KITTI, DXPR achieves near-perfect Recall@1 on most sequences and outperforms prior CMPR baselines. On Boreas, DXPR achieves intra-sequence performance on par with a strong single-modal baseline (DINOv2-SALAD), while showing clear improvements in the more challenging inter-sequence setting. Compared with RangeBEV, our method consistently performs better in both intra- and inter-sequence evaluations, demonstrating robustness under diverse seasonal and illumination changes.

Self-supervised monocular depth estimation with hybrid CNN-transformer architecture using skip attention mechanism

A transformer-based hybrid architecture for monocular depth estimation using local-global feature fusion

The Evolution of Monocular Depth Estimation:From Spatial Regression to GenerativeFoundations and the Reliability Gaps

Monocular Depth Estimation (MDE) is one of the most rigorously studied problems in modern computer vision, yet it is fundamentally ill-posed. Recovering absolute three-dimensional geometry from a single two-dimensional projection is mathematically impossible without strong inductive priors. This paper presents a structurally organized review of MDE's evolution spanning from 2005 to 2026. We trace the trajectory from handcrafted Markov random fields to brute-force pixel-wise regression with convolutional neural networks (CNNs), and then to photometric self-supervision, which liberated the field from expensive LiDAR sensor suites. We further analyzed the vision transformers (ViTs) and generative diffusion priors have achieved unprecedented zero-shot metric generalization - with models such as Metric3D v2 attaining an Absolute Relative Error (AbsRel) as low as 0.039 on the KITTI benchmark. We also show that the standard self supervised baseline Monodepth2 degrades catastrophically to an AbsRel of 1.185 under nighttime conditions in NuScenes-Night dataset, which is a massive performance collapse from its clear-weather baseline. Physical-prior models such as PhysDepth recover this to 0.118 AbsRel by embedding Rayleigh scattering theory directly into the network. At the efficiency frontier, architectures such as LEDepth achieve 0.101 AbsRel at 5.7 ms inference time with only 3.1M parameters. We conclude that the future of depth estimation lies in embedding rigid physical and spatiotemporal priors into foundational latent spaces to ensure unyielding reliability in the physical world.

MACalib-Net: Spatiotemporal multi-attention cooperative network for LiDAR-camera extrinsic calibration

Extrinsic calibration accuracy is a critical bottleneck for LiDAR-camera fusion in autonomous driving. To address motion dynamics and cross-modal disparities, this paper proposes MACalib-Net, a spatiotemporal multi-attention cooperative network for LiDAR-camera extrinsic calibration. The proposed method introduces a hierarchical Local-Global Spatial Attention mechanism, which combines local window-based geometric extraction with representative-based global context modeling for fine-grained spatial feature alignment. Furthermore, a Temporal Transformer is introduced to explicitly model inter-frame motion dependencies. Experimental results on the KITTI odometry dataset demonstrate that MACalib-Net achieves an average rotation error of 0.074° and a translation error of 0.682 cm, outperforming state-of-the-art methods in translation accuracy. Robust generalization is also demonstrated on non-repetitive scanning solid-state LiDARs (Livox Avia). Real-platform validation on a custom-built 32-line LiDAR and monocular camera system further confirms the method's deployability, achieving rotation and translation errors of approximately 0.1° and 1 cm within a two-stage calibration pipeline. The code is available at https://github.com/isjhli/MACalibNet.

Vernata: Self-Supervised Learning of LiDAR Point Representations

Vernata is introduced, consisting of three extensions: sparse view augmentation to improve robustness against varying point densities, a memory bank mechanism to stabilize resource-constrained training, and cross-modal distillation utilizing dense, high-resolution 2D image features to enable fine-grained semantic guidance.