Pyramid Feature Attention Network for Monocular Depth Prediction: Improving Accuracy and Spatial Details
Pyramid Feature Attention Network (PFANet) enhances high-level context features and low-level spatial features for accurate monocular depth prediction.
DepthFM: Fast Monocular Depth Estimation with Flow Matching
Efficient monocular depth estimation using flow matching for fast and accurate results.
Unsupervised Training for Metric Monocular Road-Scene Depth Estimation
StableCamH enables unsupervised training of monocular depth networks to learn absolute scale and metric accuracy using object size priors.
DepthFM: Fast Monocular Depth Estimation with Flow Matching
直接マッピングを使用した高速な単眼深度推定モデルの提案とその効果的な汎化能力に焦点を当てる。
Leveraging Near-Field Lighting for Monocular Depth Estimation in Endoscopy Videos
Utilizing near-field lighting for improved monocular depth estimation in endoscopy videos.
Self-supervised Indoor Monocular Depth Estimation Framework: F2Depth
F2Depth proposes a self-supervised indoor monocular depth estimation framework using optical flow consistency and feature map synthesis.
ECoDepth: Conditioning Diffusion Models for Monocular Depth Estimation
Using ViT embeddings improves monocular depth estimation.
FlowDepth: Self-Supervised Monocular Depth Estimation with Dynamic Motion Flow Decoupling and Depth-Cue-Aware Blurring
FlowDepth proposes a novel self-supervised multi-frame monocular depth estimation framework that decouples dynamic motion flow, applies depth-cue-aware blurring, and introduces a cost-volume sparse loss to address the mismatch problem, unfairness in photometric errors, and depth uncertainty in low-texture regions.
Leveraging Diffusion-Based Image Generators for Robust Monocular Depth Estimation
Marigold, a diffusion model-based method, can effectively leverage the rich visual priors captured in modern generative image models to achieve state-of-the-art performance in zero-shot monocular depth estimation across diverse real-world scenes.
Virtually Enriched NYU Depth V2 Dataset for Improving Monocular Depth Estimation
Virtually augmenting the NYU Depth V2 dataset with randomly generated 3D objects improves the performance and generalization of deep neural networks for monocular depth estimation.
